Information Systems 5

Starting with the first row of Figure 3-2, business intelligence can be used just for informing. For example, retail grocery store managers can use a BI system to see which products are selling quickly. At the time of the analysis, they may not have any particular purpose in mind but are just browsing the BI results for some future, unspecified purpose. They just want to know “how we’re doing.”

See what a typical workday would look like for someone who manages data and analytics in the Career Guide.

Moving down a row in Figure 3-2, some managers use BI systems for decision making. Managers could use a BI system or its user data to determine the location of the closest retail store to each user. They could then create customized marketing plans for each store, targeting the products those specific customers buy most often. So, for example, a grocery store chain may market expensive lobster tails at one location (the more affluent part of town) and less expensive hamburgers in another part of town. Profits might go up, and waste would probably go down. (By the way, some authors define BI systems as supporting decision making only, in which case they use the older term decision support systems as a synonym for decision-making BI systems. We take the broader view here to include all four of the tasks in Figure 3-2 and will avoid the term decision support systems.) Problem solving is the next category of business intelligence use. Again, a problem is a perceived difference between what is and what ought to be. Business intelligence can be used for both sides of that definition: determining what is as well as what should be. If revenue is below expectations, a grocery store manager can use BI to learn what factors to change to increase sales and reduce food waste. They could buy the right types of food in the correct quantities. Finally, business intelligence can be used during project management. A grocery store manager can use BI to support a project to build an in-store café. If the café is successful, it can use BI to determine which locations would be good future expansion targets. As you study Figure 3-2, recall the hierarchical nature of these tasks. Deciding requires informing; problem solving requires deciding (and informing); and project management requires problem solving (and deciding [and informing]).

Figure 3-3 shows the three primary activities in the BI process: acquire data, perform analysis, and publish results. These activities directly correspond to the BI elements in Figure 3-1. Data acquisition is the process of obtaining, cleaning, organizing, relating, and cataloging source data. Data acquired from different sources is made uniform and consistent through a process called master data management. Master data management is necessary because data from one source may not be consistently formatted with data from another source. For example, potential customer data purchased from an external data aggregator, or company that gathers and sells information from multiple sources, may not be compatible with internal operational data. Master data management ensures that all data are consistent and uniform for later analysis.

Figure 3-3: Three Primary Activities in the BI Process

We will illustrate a simple data acquisition example later in this question and discuss data acquisition in greater detail in Q3-2. BI analysis is the process of creating business intelligence. The three fundamental categories of BI analysis are reporting, data mining, and Big Data. We will describe each of the categories of BI analysis and illustrate a simple example of a reporting system later in this question and in greater detail in Q3-3 and Q3-4. We will also look at automated analysis using AI and machine learning later in Q3-5 and Q3-8. Publish results is the process of delivering business intelligence to the knowledge workers who need it. Push publishing delivers business intelligence to users without any request from the users; the BI results are delivered according to a schedule or as a result of an event or particular data condition. Pull publishing requires the user to request BI results. Publishing media include print as well as online content delivered via Web servers, specialized Web servers known as report servers, automated applications, knowledge management systems, and content management systems. We will discuss these publishing options further in Q3-4. For now, consider a simple example of the use of business intelligence.

3D printing offers the possibility for customers to print parts they need rather than order them from a retailer or distributor. One large distributor of bicycle parts wanted to stay on top of this potential change in demand and decided to investigate the possibility of selling 3D printing files for the parts rather than the parts themselves. Accordingly, it created a team to examine past sales data to determine which part designs it might sell. To do so, the company needed to identify qualifying parts and compute how much revenue potential those parts represent. To address this problem, the team obtained an extract of sales data from its IS department and stored it in Microsoft Access. It then created five criteria for parts that might quality for this new program. Specifically, it looked for parts that were:

1. Provided by certain vendors (starting with just a few vendors that had already agreed to make part design files available for sale)

2. Purchased by larger customers (individuals and small companies would be unlikely to have 3D printers or the expertise needed to use them)

3. Frequently ordered (popular products)

4. Ordered in small quantities (3D printing is not suited for mass production)

5. Simple in design (easier to 3D print)

The team knew that the fifth criterion would be difficult to evaluate because the company doesn’t store data on part complexity per se. After some discussion, the team decided to use part weight and price as surrogates for simplicity, operating under the assumption that “If it doesn’t weigh very much or cost very much, it probably isn’t complex.” At least, the team decided to start that way and find out. Accordingly, the team asked the IS department to include part weight in the extract.

A variety of data sources can be used to make decisions. What happens when these decisions include recommendations for loan approvals? The Ethics Guide considers these questions.

Acquire Data As shown in Figure 3-3, acquiring data is the first step in the BI process. In response to the team’s request for data, the IS department extracted operational data to produce the table shown in Figure 3-4. This table is a combination of data from the Sales table (CustomerName, Contact, Title, Bill Year, Number Orders, Units, Revenue, Source, PartNumber) and the Part table (PartNumber, Shipping Weight, Vendor) for select vendors willing to release 3D part design files.

Figure 3-4: Sample Extracted Data

Source: Windows 10, Microsoft Corporation.

As team members examined this data, they concluded they had what they needed and actually wouldn’t need all of the data columns in the table. Notice there are some missing and questionable values. Numerous rows have missing values of Contact and Title, and some rows have a value of zero for Units. The missing contact data and title data weren’t a problem. But the values of zero units might be problematic. At some point, the team might need to investigate what these values mean and possibly correct the data or remove those rows from the analysis. In the immediate term, however, the team decided to proceed even with these incorrect values. Such problematic data is common in data extracts.

Data breaches involving sensitive data are becoming more common. See the Security Guide for more information.

Analyze Data The data in Figure 3-4 has been filtered for the team’s first criterion to consider parts only from particular vendors. For their next criterion, team members needed to decide how to identify large customers. To do so, they created a query that sums the revenue, units, and average price for each customer. Looking at the query results in Figure 3-5, team members decided to consider only customers having more than 0,000 in total revenue; they created a query having just those customers and named that query Big Customers.

Figure 3-5: Customer Summary

Source: Windows 10, Microsoft Corporation

Next, team members discussed what they meant by frequent purchase and decided to include items ordered an average of once a week or roughly 50 times per year. They set that criterion for Number Orders in the query to select only parts that were ordered in small quantities. They first created a column that computes average order size (Units/[Number Orders]) and then set a criterion on that expression that the average must be less than 2.5. Their last two criteria were that the part be relatively inexpensive and that it be lightweight. They decided to select parts with a unit price (computed as Revenue/Units) less than 100 and a shipping weight less than 5 pounds. The results of this query are shown in Figure 3-6. Of all the parts that the company sells, these 12 fit the criteria that the team created. The next question was how much revenue potential these parts represent. Accordingly, the team created a query that connected the selected parts with their past sales data. The results are shown in Figure 3-7.

Figure 3-6: Qualifying Parts Query Results

Source: Windows 10, Microsoft Corporation

Figure 3-7: Sales History for Selected Parts

Source: Windows 10, Microsoft Corporation

Publish Results Publish results is the last activity in the BI process shown in Figure 3-3. In some cases, this means placing BI results on servers for publication to knowledge workers over the Internet or other networks. In other cases, it means making the results available via a Web service for use by other applications. In still other cases, it means creating PDFs or PowerPoint presentations for communicating to colleagues or management. In this case, the team reported these results to management in a team meeting. Judging just by the results in Figure 3-7, there seems to be little revenue potential in selling designs for these parts. The company would earn minimal revenue from the parts themselves; the designs would have to be priced considerably lower, and that would mean almost no revenue. In spite of the low revenue potential, the company might still decide to offer 3D designs to customers. It might decide to give the designs away as a gesture of goodwill to its customers; this analysis indicates it will be sacrificing little revenue to do so. Or it might do it as a PR move intended to show that it’s on top of the latest manufacturing technology. Or it might decide to postpone consideration of 3D printing because it doesn’t see that many customers ordering the qualifying parts. Of course, there is the possibility that the team members chose the wrong criteria. If they have time, it might be worthwhile to change their criteria and repeat the analysis. Such a course is a slippery slope, however. They might find themselves changing criteria until they obtain a result they want, which yields a very biased study. This possibility points again to the importance of the human component of an IS. The hardware, software, data, and query-generation procedures are of little value if the decisions that the team made when setting and possibly revising criteria are poor. Business intelligence is only as intelligent as the people creating it! With this example in mind, we will now consider each of the activities in Figure 3-3 in greater detail.

Q3-2 How Do Organizations Use Data Warehouses and Data Marts to Acquire Data?

Although it is possible to create basic reports and perform simple analyses from operational data, this course is not usually recommended. For reasons of security and control, IS professionals do not want data analysts processing operational data. If an analyst makes an error, that error could cause a serious disruption in the company’s operations. Also, operational data is structured for fast and reliable transaction processing. It is seldom structured in a way that readily supports BI analysis. Finally, BI analyses can require considerable processing; placing BI applications on operational servers can dramatically reduce system performance. For these reasons, most organizations extract operational data for BI processing. For small organizations, the extraction may be as simple as an Access database. Larger organizations, however, typically create and staff a group of people who manage and run a data warehouse, which is a facility for managing an organization’s BI data. The functions of a data warehouse are to:

· Obtain data

· Cleanse data

· Organize and relate data

· Catalog data

Figure 3-8 shows the components of a data warehouse. Programs read operational and other data, and extract, clean, and prepare that data for BI processing. The prepared data is stored in a data warehouse database. Data warehouses include data that is purchased from outside sources. The purchase of data about organizations (e.g., historical financial data) is not unusual or particularly concerning from a privacy standpoint. However, some companies choose to buy personal consumer data (e.g., marital status) from data vendors such as Acxiom Corporation. Figure 3-9 lists some of the consumer data that can be readily purchased. An amazing (and, from a privacy standpoint, frightening) amount of data is available.

Figure 3-8: Components of a Data Warehouse

Figure 3-9: Examples of Consumer Data That Can Be Purchased

Metadata concerning the data–its source, its format, its assumptions and constraints, and other facts about the data–is kept in a data warehouse metadata database. The data warehouse extracts and provides data to BI applications. The term business intelligence users is different from knowledge workers in Figure 3-1. BI users are generally specialists in data analysis, whereas knowledge workers are often non-specialist users of BI results. A loan approval officer at a bank is a knowledge worker but not a BI user.

Most operational and purchased data has problems that inhibit its usefulness for business intelligence. Figure 3-10 lists the major problem categories. First, although data that is critical for successful operations must be complete and accurate, marginally necessary data need not be. For example, some systems gather demographic data in the ordering process. But, because such data is not needed to fill, ship, and bill orders, its quality suffers.

Another approach to managing an organization’s BI data is to create a data lake, which is a central repository for large amounts of raw unstructured data. Data lakes are similar to data warehouses, but they are used for different purposes. Companies can maintain and use both a data warehouse and a data lake depending on their needs. Figure 3-12 shows some of the differences between data lakes and data warehouses.

Data lakes can contain more types of data than a data warehouse, and it can store them in their raw unstructured forms. Data lakes can also store real-time data from smart devices, websites, and mobile applications. Data lakes are useful for storing large amounts of data to be later used by data scientists in machine learning and deep learning (discussed later in this lesson). Analysis of data from data lakes can provide new insights that can’t be found in traditional data warehouses that are traditionally focused on reporting, trends, and answering operational questions. Data lakes also have their own set of unique problems. If data in a data lake are not managed and cataloged correctly, data may become inadvertently hidden over time. A company’s data lake may become a data swamp that stores large amounts of data that may never be used.

Q3-3 What Are Three Techniques for Processing BI Data?

Figure 3-13 summarizes the goals and characteristics of three fundamental types of BI analysis. In general, reporting analyses are used to create information about past performance, whereas data mining is used primarily for classifying and predicting. There are exceptions, but these statements are reasonable rules of thumb. The goal of Big Data analysis is to find patterns and relationships in the enormous amounts of data generated from sources like social media sites or Web server logs. As indicated, Big Data techniques can include reporting and data mining as well. Consider the characteristics of each type.

Reporting analysis is the process of sorting, grouping, summing, filtering, and formatting structured data. Structured data is data in the form of rows and columns. Most of the time structured data means tables in a relational database, but it can refer to spreadsheet data as well. A reporting application is a BI application that inputs data from one or more sources and applies reporting processes to that data to produce business intelligence. The team that analyzed parts in Q3-1 used Access to apply all five of these operations. Examine, for example, Figure 3-7. The results are sorted by Total Revenue and filtered for particular parts, sales are grouped by PartNumber, Total Orders and Total Revenue are calculated, and the calculations for Total Revenue are formatted correctly as dollar currency. Another type of report, exception reports are produced when something out of predefined bounds occurs. For example, a hospital might want an exception report showing which doctors are prescribing more than twice the amount of pain medications than the average doctor. This could help the hospital reduce the potential for patient addiction to pain medications. The simple operations mentioned previously can be used to produce complex and highly useful reports. Consider RFM analysis and online analytical processing as two prime examples. RFM Analysis RFM analysis, a technique readily implemented with basic reporting operations, is used to analyze and rank customers according to their purchasing patterns.5 RFM considers how recently (R) a customer has ordered, how frequently (F) a customer ordered, and how much money (M) the customer has spent. To produce an RFM score, the RFM reporting tool first sorts customer purchase records by the date of their most recent (R) purchase. In a common form of this analysis, the tool then divides the customers into five groups and gives customers in each group a score of 5 to 1. The 20 percent of the customers having the most recent orders are given an R score of 5, the 20 percent of the customers having the next most recent orders are given an R score of 4, and so forth, down to the last 20 percent, who are given an R score of 1. The tool then re-sorts the customers on the basis of how frequently (F) they order and assigns values in the same way. Finally, the tool sorts the customers again according to the amount of money (M) spent on their orders and assigns similar rankings. Figure 3-14 shows sample RFM results for Big 7 Sports (R=5, F=5, and M=3). Big 7 sports has ordered recently and orders frequently. But its M score of 3 indicates that it does not order the most expensive goods. From these scores, the sales team can conclude that Big 7 Sports is a good, regular customer and that it should attempt to up-sell more expensive goods to Big 7 Sports. No one on the sales team should even think about the third customer, Miami Municipal. This company has not ordered for some time; it did not order frequently; and, when it did order, it bought the least expensive items and not many of them. Let Miami Municipal go to the competition; the loss will be minimal.

Online Analytical Processing (OLAP) Online analytical processing (OLAP), a second type of reporting application, is more generic than RFM. OLAP provides the ability to sum, count, average, and perform other simple arithmetic operations on groups of data. The defining characteristic of OLAP reports is that they are dynamic. The viewer of the report can change the report’s format–hence the term online. An OLAP report has measures and dimensions. A measure is the data item of interest. It is the item that is to be summed or averaged or otherwise processed in the OLAP report. Total sales, average sales, and average cost are examples of measures. A dimension is a characteristic of a measure. Purchase date, customer type, customer location, and sales region are all examples of dimensions. Figure 3-15 shows a typical OLAP report. Here, the measure is Sum of store_sales, and the dimensions are Product Family (rows) and Store Type (columns). This report shows how net store sales vary by product family and store type. Stores of type Supermarket sold a net of $60,259 worth of non-consumable goods, for example.

Figure 3-15: Example Grocery Sales OLAP Report

Source: Windows 10, Microsoft Corporation.

A presentation like that in Figure 3-15 is often called an OLAP cube or sometimes simply a cube. The reason for this term is that some software products show these displays using three axes, like a cube in geometry. The origin of the term is unimportant here, however. Just know that an OLAP cube and an OLAP report are the same thing. As stated earlier, the distinguishing characteristic of an OLAP report is that the user can alter the format of the report. Figure 3-16 shows such an alteration. Here, the user added another dimension, Store Country (nested row under Product Family) and Store State (nested row under Store Country), to the horizontal display. Product-family sales are now broken out by store location. Observe that the sample data only includes stores in the United States and only in the western states of California, Oregon, and Washington.

Figure 3-16: Example of Expanded Grocery Sales OLAP Report

Source: Windows 10, Microsoft Corporation.

With an OLAP report, it is possible to drill down into the data. This term means to further divide the data into more detail. In Figure 3-17, for example, the user has drilled down into the stores located in California; the OLAP report now shows sales data for the four cities in California that have stores.

Figure 3-17: Example of Drilling Down into Expanded Grocery Sales OLAP Report

Source: Windows 10, Microsoft Corporation.

Notice another difference between Figures 3-16 and 3-17. The user has not only drilled down, but she has also changed the order of the dimensions. Figure 3-16 shows Product Family and then store location nested within Product Family. Figure 3-17 shows store location and then Product Family nested within store location. Both displays are valid and useful, depending on the user’s perspective. A product manager might like to see product families first and then store location data. A sales manager might like to see store locations first and then product data. OLAP reports provide both perspectives, and the user can switch between them while viewing the report.

Data mining is the application of statistical techniques to find patterns and relationships among data for classification and prediction. As shown in Figure 3-18, data mining resulted from a convergence of disciplines, including artificial intelligence and machine learning.

Figure 3-18: Source Disciplines of Data Mining

Most data mining techniques are sophisticated, and many are difficult to use well. Such techniques are valuable to organizations, however, and some business professionals, especially those in finance and marketing, have become expert in their use. Today, in fact, there are many interesting and rewarding careers for business professionals who are knowledgeable about data mining techniques. In an effort to make finding patterns and relationships among data more user-friendly, processes have been developed to allow users to visually analyze and explore data. This process is called data discovery. Data visualization, or the graphical representation of data, allows users to quickly understand complex data. Data discovery tools, like data visualization, are increasing in popularity because of their usefulness. However, data discovery tools may miss meaningful patterns or correlations that would be found by data mining techniques. Data mining techniques fall into two broad categories: unsupervised and supervised. We explain both types in the following sections. Unsupervised Data Mining With unsupervised data mining, analysts do not create a model or hypothesis before running the analysis. Instead, they apply a data mining application to the data and observe the results. With this method, analysts create hypotheses after the analysis, in order to explain the patterns found. One common unsupervised technique is cluster analysis. With it, statistical techniques identify groups of entities that have similar characteristics. A common use for cluster analysis is to find groups of similar customers from customer order and demographic data. For example, suppose a cluster analysis finds two very different customer groups: One group has an average age of 33, owns four Android phones and three iPads, has an expensive home entertainment system, drives a Lexus SUV, and tends to buy expensive children’s play equipment. The second group has an average age of 64, owns Arizona vacation property, plays golf, and buys expensive wines. Suppose the analysis also finds that both groups buy designer children’s clothing. These findings are obtained solely by data analysis. There is no prior model about the patterns and relationships that exist. It is up to the analyst to form hypotheses, after the fact, to explain why two such different groups are both buying designer children’s clothes. Supervised Data Mining With supervised data mining, data miners develop a model prior to the analysis and apply statistical techniques to data to estimate parameters of the model. For example, suppose marketing experts in a communications company believe that cell phone usage on weekends is determined by the age of the customer and the number of months the customer has had the cell phone account. A data mining analyst would then run an analysis that estimates the effect of customer and account age. One such analysis, which measures the effect of a set of variables on another variable, is called a regression analysis. A sample result for the cell phone example is: CellphoneWeekendMinutes = 12 + (17.5 * CustomerAge) + (23.7 * NumberMonthsOfAccount) Using this equation, analysts can predict the number of minutes of weekend cell phone use by summing 12, plus 17.5 times the customer’s age, plus 23.7 times the number of months of the account. As you will learn in your statistics classes, considerable skill is required to interpret the quality of such a model. The regression tool will create an equation, such as the one shown. Whether that equation is a good predictor of future cell phone usage depends on statistical factors, such as t values, confidence intervals, and related statistical techniques. Identifying Changes in Purchasing Patterns Most students are aware that business intelligence is used to predict purchasing patterns. Amazon made the phrase “Customers who bought . . . also bought” famous; when we buy something today, we expect the e-commerce application to suggest what else we might want. More interesting, however, is identifying changes in purchasing patterns. Retailers know that important life events cause customers to change what they buy and, for a short interval, to form new loyalties to new brands. Thus, when people start their first professional job, get married, have a baby, or retire, retailers want to know. Before BI, stores would watch the local newspapers for graduation, marriage, and baby announcements and send ads in response. That was a slow, labor-intensive, and expensive process. Target wanted to get ahead of the newspapers and in 2002 began a project to use purchasing patterns to determine that someone was pregnant. By applying business intelligence techniques to its sales data, Target was able to identify a purchasing pattern of lotions, vitamins, and other products that reliably predicts pregnancy. When Target observed that purchasing pattern, it sent ads for diapers and other baby-related products to those customers. Its program worked–too well for one teenager who had told no one she was pregnant. When she began receiving ads for baby items, her father complained to the manager of the local Target store, who apologized. It was the father’s turn to apologize when he learned that his daughter was, indeed, pregnant.

Big Data (also spelled BigData) is a term used to describe data collections that are characterized by huge volume, rapid velocity, and great variety. Considering volume, Big Data refers to data sets that are at least a petabyte in size, and usually larger. A data set containing all Google searches in the United States on a given day is Big Data in size. Additionally, Big Data has high velocity, meaning that it is generated rapidly. (If you know physics, you know that speed would be a more accurate term, but speed doesn’t start with a v, and the vvv description has become a common way to describe Big Data.) The Google search data for a given day is generated in, well, just a day. In the past, months or years would have been required to generate so much data. Finally, Big Data is varied. Big Data may have structured data, but it also may have free-form text, dozens of different formats of Web server and database log files, streams of data about user responses to page content, and possibly graphics, audio, and video files. MapReduce Because Big Data is huge, fast, and varied, it cannot be processed using traditional techniques. MapReduce is a technique for harnessing the power of thousands of computers working in parallel. The basic idea is that the Big Data collection is broken into pieces, and hundreds or thousands of independent processors search these pieces for something of interest. That process is referred to as the Map phase. In Figure 3-19, for example, a data set having the logs of Google searches is broken into pieces, and each independent processor is instructed to search for and count search keywords. Figure 3-19, of course, shows just a small portion of the data; here you can see just the keywords that begin with H.

Figure 3-19: MapReduce Processing Summary

Figure 3-21 lists four server alternatives for BI publishing. Static reports are BI documents that are fixed at the time of creation and do not change. A printed sales analysis is an example of a static report. In the BI context, most static reports are published as PDF documents.

Dynamic reports are BI documents that are updated at the time they are requested. A sales report that is current at the time the user accessed it on a Web server is a dynamic report. In almost all cases, publishing a dynamic report requires the BI application to access a database or other data source at the time the report is delivered to the user. Pull options for each of the servers in Figure 3-21 are the same. The user goes to the site, clicks a link (or opens an email), and obtains the report. Because these options are the same for all four server types, they are not shown in Figure 3-21. Push options vary by server type. For email or collaboration tools, push is manual; someone–say, a manager, an expert, or an administrator–creates an email with the report as an attachment (or URL to the collaboration tool) and sends it to the users known to be interested in that report. For Web servers and SharePoint–which you will learn about in Lesson 7–users can create alerts and RSS feeds to have the server push content to them when the content is created or changed, with the expiration of a given amount of time, or at particular intervals. SharePoint workflows can also push content. A BI server extends alert/RSS functionality to support user subscriptions, which are user requests for particular BI results on a particular schedule or in response to particular events. For example, a user can subscribe to a daily sales report, requesting that it be delivered each morning. Or the user might request that RFM analyses be delivered whenever a new result is posted on the server, or a sales manager might subscribe to receive a sales report whenever sales in his region exceed $1M during the week. We explain the two major functions of a BI server in the next section. The skills needed to create a publishing application are either low or high. For static content, little skill is needed. The BI author creates the content, and the publisher (usually the same person) attaches it to an email or puts it on the Web or a SharePoint site, and that’s it. Publishing dynamic BI is more difficult; it requires the publisher to set up database access when documents are consumed. In the case of a Web server, the publisher will need to develop or have a programmer write code for this purpose. In the case of SharePoint and BI servers, program code is not necessarily needed, but dynamic data connections need to be created, and this task is not for the technically faint of heart. You’ll need knowledge beyond the scope of this class to develop dynamic BI solutions. You should be able to do this, however, if you take a few more IS courses or major in IS.

A BI server is a Web server application that is purpose-built for the publishing of business intelligence. The Microsoft SQL Server Report manager (part of Microsoft SQL Server Reporting Services) is the most popular such product today, but there are other products as well. BI servers provide two major functions: management and delivery. The management function maintains metadata about the authorized allocation of BI results to users. The BI server tracks what results are available, what users are authorized to view those results, and the schedule upon which the results are provided to the authorized users. It adjusts allocations as available results change and users come and go. As shown in Figure 3-22, all management data needed by any of the BI servers is stored in metadata. The amount and complexity of such data depends, of course, on the functionality of the BI server.

Figure 3-22: Elements of a BI System

BI servers use metadata to determine what results to send to which users and, possibly, on which schedule. Today, the expectation is that BI results can be delivered to “any” device. In practice, any is interpreted to mean computers, smartphones, tablets, applications such as Microsoft Office, and standardized Web applications.

Nothing is more frustrating for a manager to contemplate than the situation in which one employee struggles with a problem that another employee knows how to solve easily. Or to learn of a customer who returns a large order because the customer could not perform a basic operation with the product that many employees (and other customers) can readily perform. Even worse, someone in the customer’s organization may know how to use the product, but the people who bought it didn’t know that. Knowledge management (KM) is the process of creating value from intellectual capital and sharing that knowledge with employees, managers, suppliers, customers, and others who need that capital. The goal of knowledge management is to prevent the kinds of problems just described. Knowledge management was done before social media. Before we turn to specific technologies, however, consider the overall goals and benefits of KM. KM benefits organizations in two fundamental ways:

· It improves process quality.

· It increases team strength.

Process quality is measured by effectiveness and efficiency, and knowledge management can improve both. KM enables employees to share knowledge with each other and with customers and other partners. By doing so, it enables the employees in the organization to better achieve the organization’s strategy. At the same time, sharing knowledge enables employees to solve problems more quickly and to otherwise accomplish work with less time and other resources, hence improving process efficiency.7 Additionally, successful teams not only accomplish their assigned tasks, but they also grow in capability, both as a team and as individuals. By sharing knowledge, team members learn from one another, avoid making repetitive mistakes, and grow as business professionals. For example, consider the help desk at any organization, say, one that provides support for electronic components like iPhones. When a user has a problem with an iPhone, he or she might contact Apple support for help. The customer service department has, collectively, seen just about any problem that can ever occur with an iPhone. The organization as a whole knows how to solve the user’s problem. However, that is no guarantee that a particular support representative knows how to solve that problem. The goal of KM is to enable employees to be able to use knowledge possessed collectively by people in the organization. By doing so, both process quality and team capability improve.

Two human factors inhibit knowledge sharing in organizations. The first is that employees can be reluctant to exhibit their ignorance. Out of fear of appearing incompetent, employees may not submit entries to blogs or discussion groups. Such reluctance can sometimes be reduced by the attitude and posture of managers. One strategy for employees in this situation is to provide private media that can be accessed only by a smaller group of people who have an interest in a specific problem. Members of that smaller group can then discuss the issue in a less-inhibiting forum. The other inhibiting human factor is employee competition. “Look,” says the top salesperson. “I earn a substantial bonus from being the top salesperson. Why would I want to share my sales techniques with others? I’d just be strengthening my competition.” This understandable perspective may not be changeable. A KM application may be ill-suited to a competitive group. Or the company may be able to restructure rewards and incentives to foster sharing of ideas among employees (e.g., giving a bonus to the group that develops the best idea). If these two factors are limiting knowledge sharing, strong management endorsement can be effective, especially if that endorsement is followed by strong positive feedback. Overcoming employee resistance can be difficult, but remember, “Nothing wrong with praise or cash . . .  especially cash.”

One form of knowledge management concerns knowledge that is encoded in documents. Content management systems (CMS) are information systems that support the management and delivery of documents including reports, Web pages, and other expressions of employee knowledge. Typical users of content management systems are companies that sell complicated products and want to share their knowledge of those products with employees and customers. Someone at Toyota, for example, knows how to change the timing belt on the four-cylinder 2022 Toyota Camry. Toyota wants to share that knowledge with car owners, mechanics, and Toyota employees.

Content management systems face serious challenges. First, most content databases are huge; some have thousands of individual documents, pages, and graphics. Second, CMS content is dynamic. Imagine the frequency of Web page changes at Apple or Google or Amazon that must occur each day! Another complication for content management systems is that documents do not exist in isolation from each other. Documents refer to one another, and when one changes, others must change as well. To manage these connections, content management systems must maintain linkages among documents so that content dependencies are known and used to maintain document consistency. A fourth complication is that document contents are perishable. Documents become obsolete and need to be altered, removed, or replaced. For example, Microsoft’s new release of Microsoft Office 2019 likely affects thousands of internal documents, external facing pages, blogs, etc. All of that has to be changed in a matter of hours. Finally, content is provided in many languages. 3M has tens of thousands of products, some of which are harmful when used improperly. 3M must publish product safety data for all such products in several dozen languages. Every document, in whatever language it was authored, must be translated into all languages before it can be published on 3M’s site. And when one document changes, all of the translated versions must change as well.

Q3-5 Why Is Artificial Intelligence (AI) Important?

Artificial intelligence (AI) is the ability of a machine to simulate human abilities such as vision, communication, recognition, learning, and decision making in order to achieve a goal. Organizations hope to use AI to increase the automation, or the process of making systems operate without human intervention, of mundane tasks typically done by humans. For example, a professor at Georgia Tech trained IBM’s AI, named Watson, to be a teaching assistant in his class. The AI was responsible for answering a portion of nearly 10,000 student questions. None of the students realized “Jill Watson” was a machine until the end of the semester when the professor identified their cybernetic classmate.8 Tim Cook, CEO of Apple, recently talked about the company’s efforts to make a self-driving car code-named Project Titan. Cook said Apple is focusing on creating the “mother of all AI projects” to make the brain of an autonomous system.9 This AI could then be used in many different autonomous Apple systems beyond self-driving cars. These could include robots, commercial drones, agricultural systems, military platforms, surgical systems, and other smart devices. Apple isn’t alone in focusing on AI development. According to a 2020 survey by PricewaterhouseCoopers, 20 percent of U.S. CEOs plan to deploy AI company wide, and an additional 27 percent have already implemented AI in multiple areas within their organizations.10 A recent report by Grand View Research suggests that the AI market could be valued at $390 billion by 2025.11 The resulting efficiency gains and cost reductions over the subsequent 10 years could be as large as $14 to $33 trillion. That’s an enormous economic impact in a short period of time. To put that number in perspective, the entire annual gross domestic product (GDP) for the U.S. is about $18 trillion.

The potential benefits of AI are mind-boggling. Studies throw out many statistics; for example, there’s a 50 percent chance AI will outperform humans in all tasks in 45 years, or there’s an 83 percent chance AI will take over all jobs paying less than $20 an hour within a few years.12 But are these studies accurate? Historically, studies such as these have not been accurate. But they might actually turn out to be correct in the near future. In the 1950s, 1980s, and late 1990s AI research went through several waves of innovation. During each wave, there was a lot of discussion about how AI would revolutionize the world. Some people got excited about it, and some got nervous about it. In the end, it led to a bunch of great science fiction movies. AI research was slow and incremental during each of these periods. Unfortunately, AI started to develop a bad reputation. In fact, the term AI was so loaded with historical false starts that when IBM started developing its Watson, it used the term cognitive computing rather than AI. But significant innovations in the past couple of decades have advanced AI to the point that it’s now starting to be successful. Figure 3-23 shows the six main forces that have helped advance AI in recent years.

Figure 3-23: Forces Driving AI Innovation

First, computing power has been increasing exponentially for several decades (Moore’s Law), while earlier waves of AI lacked the necessary computing power. The computers during those time periods were just not fast enough to process the data necessary to make AI work. Just like an airplane during takeoff, AI must reach critical speed before it can get off the ground. IBM’s Watson, for example, operates at about 80 teraflops per second. Second, the availability of large data sets has advanced to the point that AI is viable technology. AI applications require a surprisingly large amount of data in order to learn, represent knowledge, and process natural language. Watson, for example, can read 800 million pages per second. AI applications need both large amounts of rich data and the accompanying processing power to sift through it. Third, cloud computing has advanced AI development because it has made scalable resources available at very low costs. AI development can now be done by more people at a much lower cost. The cloud allows developers to access existing AI applications through online interfaces. IBM offers developers access to a variety of online AI applications including Conversation, Discovery, Document Conversion, Language Translator, Tone Analyzer, Speech to Text, and Visual Recognition. Fourth, the rapid increase in network connected smart devices is producing vast amounts of data for AI applications. These devices are more than just desktops, laptops, and smartphones; they also include a variety of network-enabled optical, motion, temperature, audio, and magnetic sensors. There are thousands of IoT devices that can provide the data AI applications use to learn. AI applications are also used by these connected devices. Fifth, fundamental breakthroughs in AI techniques have made AI useful for a variety of tasks. For example, in 2006, AI researcher Geoffrey Hinton developed a method for simulating multiple layers of artificial neural networks rather than just a single layer.13 A neural network is a computing system modeled after the human brain that is used to predict values and make classifications. This multilayered neural network technique was applied to learning tasks and is now commonly known as deep learning. Deep learning has greatly increased the accuracy and practical usefulness of AI. Finally, recent advances in AI have been driven by the demand for applications that solve practical problems. Figure 3-24 shows a few examples of practical AI applications produced by a handful of the largest tech companies. These are not academic theories or artificial simulations; these are applications focused on solving real-world problems for organizations. There are many more applications than those shown in Figure 3-24. In fact, Google has more than 1,000 working AI projects, and Apple and Microsoft are integrating AI into all of their hardware and software.

These six forces have driven innovations in AI to the point that AI is now becoming widely used, and it’s becoming a core part of many tech companies’ strategic advantage. AI has already been widely adopted by all of the major tech companies, and users are interacting with it (sometimes unknowingly) on a daily basis. It’s also being used to create innovative new products by smaller companies. Traditional organizations are starting to see the value of AI, too.

Knowledge Check

Q3-6 How Will Artificial Intelligence and Automation Affect Organizations?

Artificial intelligence sounds like a great innovation–it will reduce costs, increase productivity, create new services, find unique solutions to age-old problems, and enable an army of smart devices with new capabilities. But, as a business leader, you need to understand the broader implications of using AI and automated machines. You need to understand how this new wave of AI is going to affect you and your organization. For example, suppose you own a chain of fast-food restaurants. You’re considering buying a fully automated hamburger-making machine powered by an AI. It can make 400 custom hamburgers an hour. It does it safely, cleanly, and without any human intervention. It’s so efficient that it can replace three full-time cooks, potentially saving you $90,000 per year. It doesn’t take breaks, call in sick, steal food, or sue you for lost wages. This robotic fast-food worker may sound far-fetched, but these types of machines already exist: Both Momentum Machines and Miso Robotics make robots with these capabilities.14 Will this new AI-powered hamburger-making machine really save you money? If so, how much? And how will your human employees react to their new cybernetic coworker?

First, let’s consider the reduction in labor costs that come from using automated labor. According to the U.S. Bureau of Labor Statistics, the average employee in the United States makes $24.36 per hour.15 But that’s not the true cost. As shown in Figure 3-25, benefits add an additional 30 percent ($10.37) per hour to the true cost. So, a $24-per-hour employee actually costs at least $35 per hour. If you buy an automated system, you won’t have to pay for any of the additional benefits required with human labor. You won’t have to pay for overtime, leave, insurance, or retirement contributions.

Figure 3-25: Employee Costs per Hour (U.S.)

Source: Based on Data from Bureau of Labor Statistics

In the case of your AI-powered hamburger machine, suppose you’re replacing three cooks who each make $15.00 per hour plus $6.90 in benefits. That’s a cost of $21.90 per cook per hour, for a total cost of $65.70 per hour. An AI-powered hamburger machine pays itself off pretty quickly at that rate. Suppose the machine costs $100,000 and runs 12 hours per day. You would recoup your investment in 4.2 months. There are other productivity gains to consider beyond wages and benefits. Figure 3-26 lists a few examples of benefits and productivity gains you could realize from using an automated hamburger machine instead of a human laborer. There are benefits of human labor, but they are not as easily quantified as the benefits from automated labor. Some of the productivity gains from automated labor would vary substantially depending on the specific circumstances. For example, if your restaurants were open 24 hours per day, you would realize sizable productivity gains from an automated hamburger machine. But if the restaurants were only open 8 hours per day, the gains might be less.

A good example of an organization benefiting from an automated system is online banking. Banks that offer consumers online banking have seen both increases in productivity and decreases in costs. For example, the average cost per transaction to talk with a teller at the bank is about $4. But per-transactions costs drop to $0.17 for online transactions and $0.08 for mobile transactions.16 Consumers can access their accounts online at any time without traveling to the bank to talk with a human teller. Online banking has allowed banks to reduce their labor costs and boost profitability. Beyond productivity gains and reductions in labor costs, other factors may impact your decision to adopt the automated hamburger machine. Some of these factors aren’t pleasant to think about because people don’t want them to happen. Let’s look at one of the costlier factors–employee fraud. Managers don’t want employees to steal from the company. It’s no fun to think your workers are stealing from the company. But employee fraud happens, and it is extremely costly. The Association of Certified Fraud Examiners (ACFE) 2020 Report to the Nations estimates the median loss for employee fraud is $125,000 per incident.17 Figure 3-27 shows the type of employee fraud by frequency and median loss per incident. Not all types of fraud are the same. Some types of fraud, like financial statement fraud, don’t occur that often but have large median losses. The ACFE estimates that a typical organization loses 5 percent of its annual revenue to employee fraud. To get an idea of the size of these losses, consider that the U.S. gross domestic product (GDP) was $21 trillion in 2019. That would mean the estimated amount of employee fraud losses would be over $1 trillion per year in the United States.

Figure 3-27: Employee Fraud Frequency and Median Loss

Source: Based on Report to the Nations on Occupational Fraud and Abuse.

If you bought an automated machine, you would reduce the amount of employee fraud within your organization dramatically. An automated machine won’t steal from you. It doesn’t want to or need to. It doesn’t feel financial pressures or look for opportunities to steal. It just performs its task. It may also indirectly add 5 percent to your bottom line by reducing employee fraud. Finally, there are a few additional benefits of automated labor that you must consider. Automated machines don’t need severance packages, don’t have union overhead, can’t go on strike, don’t steal intellectual property, won’t file discrimination lawsuits, and can’t harass coworkers. With human labor, all of these add costs to your organization’s bottom line. They consume your time and energy as a manager, and they affect your organization’s ability to remain competitive.

Whenever people start talking about the impacts of AI and automation, there are typically two distinct reactions. The first reaction is from the group of people who see AI as an incredible opportunity. They get excited about the gains in productivity, profitability, and competitive advantages. They want to be on the cutting edge and be the first to implement it in their industry. The second reaction is from the group of people who see AI as a serious threat. They’re worried about their jobs. They are concerned about what happens if they get replaced by a machine or an AI bot. They worry that their years of education and training will become worthless or that they may have to change careers. These are valid concerns, and this group of people is right to be concerned. Major seismic shifts in global workforces are imminent. Jobs requiring routine physical and mental tasks are prime candidates for automation. Experts on AI and automation estimate that by 2030, 30 to 40 percent of these types of workers could be taken out of the current worldwide labor force.18 They will be replaced by AI with an IQ higher than 90 percent of the U.S. population. However, overall employment is expected to remain flat due to increased demand for workers with high digital and technical skills. As a manager, you need to understand the economic implications of these historic changes. They’re important for your organization, and they might be important to you personally.

While workers may not want their incomes to go away, they may not mind seeing certain jobs go away (assuming they can transition to a different job that still pays a good salary). Some jobs are dirty, stinky, boring, and even dangerous. AI and automation could do jobs that humans don’t really want to do. Take waste disposal, for example. Twenty years ago, there were three workers on a garbage truck. One would drive, and two would fill the truck. Today, one person drives the truck and operates the side loader arm. In another twenty years, the truck will be self-driving, and garbage collection will be fully autonomous. On a personal level, AI and autonomous machines could do other things you may not want to do. Home repairs, house cleaning, car washing, auto repairs, gardening, and cooking could all be done autonomously. Thanks to your new “gardener,” you could have fresh homegrown food every day. You wouldn’t have to buy as much food from the grocery store. Your personal AI could cook healthy meals and wash your dirty dishes. Your personal costs would go down, and your productivity would go up. AI and autonomous machines could also take care of the elderly and disabled, provide companionship, train “energetic” puppies, and teach children without becoming frustrated. Similarly, there are jobs within organizations that humans may not want to do. Manufacturing and agricultural jobs have historically been seen as monotonous, backbreaking, and low-paying. They’ve also been decreasing in the United States over the past 40 years. Automated systems may allow workers to shift from assembling and making existing products to designing and creating new products. The key will be helping workers make that shift.

Retraining and Retooling

Critics of AI and automated machines claim that adoption of these systems will lead to mass unemployment. Similar predictions of a technological apocalypse have been made before, but it hasn’t happened. For example, in the late 1990s, online shopping started to become more commonplace. Businesses weren’t sure if brick-and-mortar businesses were going to be taken over by their online competition. Were they going to have to fire their traditional employees and transition to an online presence? A period of excessive speculation in tech companies called the dot-com boom began. Investors poured money into tech startups in hopes of making billions. Most of them failed by 2001. The spectacular boom and bust of these tech companies got a lot of attention in the media. But the underlying shift in the global workforce was an equally important yet subtle change. New types of jobs were available to workers. Jobs in networking, database development, Web development, and programming paid well and were in high demand. Workers retrained for the new digital era. They sought out skills demanded by organizations that shifted their strategic priorities. A similar retooling of the existing workforce will be necessary when AI and automated machines take a more prominent role in our economy. There will be new types of jobs for humans to do. Lots of them.

How are human workers going to survive the workplace shift caused by AI? First, human workers need to develop skills that machines can’t do. Back in Lesson 1 we touched on a few nonroutine cognitive skills that help keep you competitive in a changing job market. These included abstract reasoning, systems thinking, collaboration, and experimentation. This is a good start. Humans excel in these areas. But there are a few other things that machines still can’t do. Creativity, adaptability, and new undefined problem solving are human skills that give you a competitive advantage over your synthetic counterparts. Second, lots of human workers will be needed to take care of these new machines. Even the best AIs need training. IBM’s Watson needs experts to train it so it knows if it’s getting the right answers. Human domain experts will always be needed. As a result, you’ll change jobs much more quickly than workers did in the past. Human workers will need to adapt quickly to their synthetic coworkers. AI-enabled machines will also need humans for the foreseeable future because they lack something humans have–instincts. Machines don’t have the basic internal unlearned driving forces that humans have. These forces ensured the survival of the human species. Without instincts, machines won’t last long at all. Machines don’t need or desire anything. Instincts drive humans to stay alive, procreate, improve their position, and seek the protection of others. Without these instincts, machines don’t care if they die, and they won’t replicate and won’t improve. Machines need humans to survive. They also need humans to help them to improve. Humans and machines need each other. That’s why we’ll be working together for a long time to come.

Q3-7 What Is the Goal of AI?

The goal of AI research is to create artificial general intelligence, or strong AI that can complete all of the same tasks a human can. This includes the ability to process natural language; to sense, learn, and interact with the physical world; to represent knowledge; to reason; and to plan. Most AI researchers believe we will have strong AI capabilities sometime around 2040.19 Currently, as shown in Figure 3-28, we have weak AI that is focused on completing a single specific task. There is speculation that someday we may be able to create an AI that moves beyond strong AI to create a superintelligence capable of intelligence more advanced than human intelligence. Some researchers see superintelligence as a potential threat to humans. Others disagree and argue that this level of AI is hundreds of years away.20

Figure 3-28: Evolution of AI Abilities

There is considerable disagreement about what it means to actually create an artificial intelligence. An early computer scientist named Alan Turing said a machine could be considered intelligent if a human could have a conversation with it and not be able to tell if it was a machine or a human. This standard, shown in Figure 3-29, is known as the Turing test. There are other standards for judging AI, but they’re beyond the scope of this course. Again, the overall goal of AI is to create a machine that can complete the same tasks as a human.

Figure 3-29: Turing Test But AI is much more than just the ability of a chat bot to simulate a human conversation. It’s the ability of a machine to simulate all human abilities. Consider the scope of the major research areas within AI as well as a few select sub-areas, as shown in Figure 3-30. These major areas focus on different aspects of human abilities. In fact, AI as a research area is much broader than can be shown here. AI is a general term that means different things to different people depending on their area of interest. However, the implementation of AI into real-world technology usually involves combining multiple different areas together.

Figure 3-30: Major AI Research Areas

Organizations see AI as an enabler of new technologies. Their goal is to use AI to enhance their existing products and services. For example, consider the amount of AI required for a fully autonomous self-driving car. A self-driving car will have multiple computer vision systems including GPS, gyroscopes, accelerometers, LIDAR, RADAR, 360-degree cameras, and possibly even night-vision capability. It will also have a robotics component that governs locomotion, sensing, and navigation. Future self-driving cars will automatically learn from your past transportation needs and environmental preferences (machine learning), preplan your routes, and monitor for delays (planning). You’ll probably even be able to give your car instructions by talking with it normally (natural language processing). Technology companies like Apple want to put AI into more than just self-driving cars. Their goal is to put it into all of their devices and services. They want all of their devices to be more than just “smart” (i.e., connected to the Internet); they want them to be intelligent (i.e., powered by an AI back-end system). They want all IoT devices to be AI driven. Think about the new products and services that could be created if AI was applied to areas like manufacturing, finance, medicine, cybersecurity, transportation, education, entertainment, and agriculture. Major technology companies are investing heavily in AI to make this happen. Saying No to AI Widespread adoption of AI is going to cause a lot of changes in organizations. As organizations change, so will the types of jobs they need filled. Workers are going to have to continually develop new skills for new types of jobs. They might find themselves doing work that has little to do with their formal education. People don’t like change in general. Change introduces risk, uncertainty, and loss of control. It will be tempting to say “no” to AI and automation. Governments may feel pressure to place restrictions on implementations of AI and automation to protect workers. But what will happen if you don’t automate and your competitors do? You may lose a competitive advantage if your competitor can produce a higher-quality product at one-third the price. Imagine banning online banking because it might cost tellers their jobs or banning Netflix because it might reduce the demand for labor at local brick-and-mortar video stores. That may sound silly to most people, but innovations have caused substantial shifts in the workforce. They always have. There’s a long list of jobs that once existed that don’t anymore. But there’s also a list of new jobs that were created because of technological innovations. Adopting AI and automation at an appropriate pace may be the only way to keep organizations viable. Looking at the adoption of AI more broadly, it may solve some of the financial woes of certain world economies. Costs for health care could be reduced by 0 billion annually.21 Employment costs could be reduced by trillion. World economies could also see productivity gains of at least 30 percent. Governments may not want to see their citizens forced into new types of jobs, but they may need the economic gains AI and automation represent. Saying “yes” to AI may be a painful but necessary decision.

Q3-8 How Does AI Work?

AI has become somewhat of a buzzword in the tech and business worlds. People talk about the amazing things AI can do, but they don’t really understand how it works. As a business professional, you need to have a basic understanding of how AI works. You don’t have to become an expert, but you do need to understand the ways it might be applied to solve organizational problems. This will enable you to actually create value within your organization rather than just being awed by the productivity gains AI has created in other organizations. The following example looks at the way AI can be used to solve a real-world problem–spam filtering. There are many other areas of AI that can produce applications with similarly compelling results, but this will give you an idea of what’s possible.

Now that you’ve seen a simple example of how AI works, let’s take a closer look at a more complex AI. IBM’s artificial intelligence named Watson is a question answering system that draws on several areas of AI. First, it uses natural language processing (NLP), or the ability of a computer system to understand spoken human language, to answer questions. It was designed to play against world champions on the quiz show Jeopardy!, and in 2011 it won. But so what? Why spend millions of dollars to build a great AI that can win a trivia game? Isn’t that trivial by definition? No. The trivia game was meant to be an exhibition of Watson’s ability to answer difficult questions asked by humans in a natural way. The implications of Watson’s win are profound. Watson can provide evidence-based answers for questions in fields like health care, transportation, education, social media, customer service, and security.23 For example, H&R Block is using Watson to help do taxes, and LegalMation is using Watson to help automate litigation. The list of potential applications for Watson is long. To better understand Watson’s capabilities, you need to understand the basics of how Watson works. How Does Watson Work? Figure 3-32 shows a shortened version of how Watson works. The actual DeepQA architecture is much more complex, and it is beyond the scope of this course.24 As a business manager it’s important to understand how Watson works so you can identify potential applications for this type of AI.

Figure 3-32: IBM’s Question-and-Answer Process

First, Watson acquires content from sources like dictionaries, encyclopedias, literature, reports, and databases. It extracts valuable pieces of data from these structured and semi-structured data sources. It then takes these extracted pieces of data and adds them to a corpus of knowledge, or a large set of related data and texts. During the Jeopardy! challenge, Watson used 200 million pages of content on 4 terabytes of disk space.25 Once the corpus is built, it can start answering questions. Each question that comes in goes through question analysis. Watson identifies the type of question being asked and analyzes the question itself. While this may sound overly simplistic, it’s good to remember that it takes most humans several years to learn to speak their primary language. Next, Watson generates hypotheses about what might be the right answer to the question. It searches its data for possible candidate answers. It takes the top 250 candidate answers and then filters them down to the top 100 answers. Then it goes back to its data and looks for evidence to support each candidate answer. It uses many different techniques to score each answer based on the available evidence. Finally, Watson merges all of the scored candidate answers, identifies the best possible answer, and estimates the probability that the answer is correct. And here’s the best part–it does everything from question analysis to final answer estimation in 3 seconds! That’s powerful. The Future for Watson IBM’s Watson is an amazing system. Watson can read hundreds of millions of pages per second, can interact with people all over the world at the same time, and speaks nine languages. But Watson still needs to be trained for different tasks. That might take more than we think. It took Watson 5 years to become the best at answering trivia questions. But Watson did it. And remember that Watson will live longer than any human and can get processor upgrades on demand that are twice as fast every 18 months (Moore’s Law). Twenty years from now, Watson may be doing many things we thought were reserved for human workers.

Q3-9 2031?

BI systems are widely used today. Simple systems using RFM and OLAP are easy to use and truly do add value. More complex systems like AI machine learning are starting to be used by large companies with success. But they must be correctly designed and implemented and applied to problems appropriately. Companies are already pretty savvy when it comes to using BI to effectively target customers with products they will likely buy. They use BI to know what you want to buy, when you’ll buy it, and how you’ll buy it. They can detect fraudulent credit card purchases, automatically lock your credit card, and resolve the unauthorized charges quickly. But there’s much more that BI can do beyond retail. By 2031, data storage, processing power, and network speeds will have increased exponentially. Most of the devices around you will be collecting and transmitting data. Companies will know more about you than just your purchasing habits. They will know your location throughout the day. They’ll know your sleeping patterns, exercise routines, stress levels, and food preferences. They will even know what you talk about in your home. All of these types of data can be used for beneficial outcomes. But they could also be used for malicious outcomes. Privacy will become an increasing concern as BI develops over time. There’s also the real possibility that many jobs will be automated by 2031. What happens when an AI becomes sophisticated enough to replace BI analysts? What happens when the AI has the ability to find its own data sets, perform its own analysis, make decisions, and then decide which analysis to perform next? AI will likely automate many of the routine analytical tasks we now know we want to perform. But it also has the potential to find new patterns, correlations, and insights. In the future, AIs may do as much teaching as they do learning. They may even move beyond us. Ray Kurzweil developed a concept he calls the Singularity, which is the point at which an AI becomes sophisticated enough that it can adapt and create its own software and, hence, adapt its behavior without human assistance. Apply this idea to unsupervised data mining.26 What happens when machines can direct their own data mining activities? There will be an accelerating positive feedback loop among AIs. A single AI will have more processing power than all possible human cognitive power combined. We may even have the technology to merge human intelligence with AIs and gain knowledge that we could never have comprehended before. Kurzweil predicts this could happen by 2045. By 2031 we’ll start to see that future. We’ll start to see machines as more than just things we create to augment ourselves (e.g., right now, we create cars because we can’t run fast enough or far enough). We’ll probably start seeing machines as assistants, coworkers, advisors, even friends. They may become the creators of things. They probably will be our caretakers. Hopefully, they will not be our overlords.

So What? Continuous Intelligence

Have you ever hopped in the car and started up the Waze navigation app to help you get to your next destination? Apps like Waze have become mainstream as drivers are seeking the most accurate and up-to-date driving conditions in the region. Not only does the app track all users in the area to generate drive-time estimates based on traffic flows and developing congestion, but users also contribute to drive-time forecasting by posting updates about hazards, traffic, police checkpoints, detours, and evolving weather conditions. At a foundational level, there are also thousands of volunteers (yes, they work for free!) who maintain the accuracy of Waze maps to ensure that long-term detours, new roads, and roads put out of service are all updated in the system as quickly as possible. In short, the value of the Waze service is the real-time amalgamation of information from a variety of sources (map editors, active reports by drivers, and passive monitoring by drivers’ apps as they drive around) and the generation of actionable information, in the form of navigation guidance, for each specific driver. The value proposition of this service is clearly visible as Waze reports that roughly 115 million drivers and riders take advantage of its platform. You may be wondering if it would be possible to take the concept of compiling multiple data streams in real time to generate actionable insights that could help navigate a business optimally to long-term success. The answer is that this very concept is actually an emerging business solution gaining traction right now; it has been termed continuous intelligence (CI).

Source: Panchenko Vladimir/Shutterstock

Business Intelligence or Continuous Intelligence? First, continuous intelligence sounds a lot like business intelligence (BI), a term with which you are probably quite familiar. However, there is a distinct difference between the two. BI systems are much less sophisticated in that they often require extensive intervention from users to configure, introduce, and format data streams; trigger analyses; and define relevant dashboards. Further, the types of data that can be introduced into BI systems are often limited.27 As organizations have recognized the incredible value of compiling and analyzing various data streams to yield actionable information (which is increasingly necessary and abundant in our highly competitive global economy), businesses started capturing and storing every data set that they could feasibly acquire. Over time, this data hoarding approach translated to troves of data in different formats that were very time-consuming to clean and standardize before they could be loaded into BI platforms. Recently, companies have begun to see that the power of artificial intelligence (AI) and machine learning could be harnessed to sift through these massive data sets to dynamically identify patterns, relationships, and meaningful insights with little or no human intervention. Suddenly, the burden of too much data became an incredible asset as continuous intelligence systems can “infer and harmonize” these vast data sets with little or no guidance. The movement to employ CI systems has been accelerated by the permeation of Internet of Things (IoT) devices and sensors, which help fill data storage. Furthermore, cloud-based solutions, like Amazon Web Services (AWS), can serve as dynamic repositories for rapid scaling of data storage based on fluctuations in data compilation efforts.29 This helps companies avoid having to invest in their own IT infrastructures, which may not benefit from data storage capabilities that are not always filled to peak capacity. All of these factors have translated to estimates that a majority of new business platforms will be leveraging continuous intelligence functionality by 2022. Organizations are more data-driven now than ever before. Commonplace methods of storing and analyzing data to aid in decision making are cumbersome and unsophisticated. IoT, cloud computing, machine learning, and AI are all making it possible for companies to rapidly, dynamically, and efficiently compile vast amounts of dissimilar data to generate meaningful insights. These new continuous intelligence platforms offer an exciting next step to help companies compete in an increasingly competitive global economy. Questions

1. The article references volunteers who work countless hours for free to ensure that the Waze map system is up to date. This is an example of crowdsourcing. What are other examples of crowdsourcing you may have benefited from?  Show Answer

2. Describe all of the different types of data streams that companies can compile to submit to a continuous intelligence platform for analysis.  Show Answer

3. How can continuous intelligence be used within the context of information security?  Show Answer

4. Can you identify any potential risks of an organization using a continuous intelligence platform?  Show Answer

Security Guide

Capital Data Breach What is your favorite type of movie–drama, comedy, romance, action? Some people especially love heist movies in which a ragtag team of crooks with various expertise are assembled to steal invaluable art, jewelry, or straight-up piles of cash. Like most things you see in the movies, they are quite a departure from reality. Today, many of the great heists do not target tangible riches in highly secured vaults. They’re not perpetrated by expansive teams either. Today’s heists are committed by lone actors or only a handful of collaborators. Most boring of all is that these heists are not after diamonds or jewels, but, rather, these criminals are seeking the contents of databases–at their most basic level, just 1s and 0s. Many companies have become victims of these heists. In July 2019, Capital One found out it was one of these victims.

Source: Ascannio/Shutterstock

Crime with a Capital C Many organizations today are faced with weighing the pros and cons of storing their data within their own corporate IT infrastructures or in the cloud. While storing data locally can ensure greater control and peace of mind (i.e., I can point to where my data is stored!), large-scale cloud providers can hire the best security practitioners and employ leading cutting-edge security best practices, and outsourcing the burden of hosting and securing data to cloud companies provides opportunities for cloud users to focus on other initiatives. Capital One elected to store data in the cloud with Amazon Web Services (AWS), and it paid a price. Its data was compromised by a former AWS employee who was able to download data constituting over 100 million credit card applications. Specifically, this data set included 140,000 social security numbers and 80,000 bank account numbers. (It also impacted 6 million individuals in Canada.)30 An analysis of the breach revealed that the point of exploitation to access the Capital One data set in the cloud was a misconfigured web application firewall.31 (This type of exploitation is technically referred to as a server side forgery request [SSRF].) Luckily, no credit card numbers or login credentials were compromised. After the attack, the perpetrator left a bread-crumb trail online referring to their hack; they were ultimately charged with one count of computer fraud and abuse. Even though it is not believed that the attacker widely distributed the compromised files or used stolen data to commit fraud, Capital One estimated that the breach will likely cost it $150 million.32 Financial institutions spend incredible amounts of money to safeguard their systems and data from these types of breaches. For example, the chief of JPMorgan Chase reported a security budget exceeding $600 million, and Bank of America apparently puts no ceiling on security expenditures. These security war chests are used to hire top talent, implement extremely sophisticated hardware solutions (e.g., firewalls and intrusion detection systems), implement and enforce organizational security policies, and train users on how to abide by these policies. While limitless funds may seem like a good thing in the war against nefarious digital actors, do you think it is possible that too much security can be a bad thing? Enough Is Enough The Capital One story is just one more in a string of high-profile cybersecurity incidents. In fact, it is not uncommon to watch the nightly news or open a newspaper and learn about a large organization being hacked. Ironically, it is possible that people hearing about data breaches regularly may feed into the occurrence of breaches. Information security researchers have been investigating the notion of information security fatigue, which they describe as a state of mind that occurs when a person has become tired and disillusioned with security-related initiatives. In other words, a person who regularly hears about cyber incidents may assume that they are so common as to be inevitable and that there is little that each individual person can do to try to prevent a breach from occurring. Sadly, this realization can cause people to give up on following security best practices and become a part of the problem rather than being a part of the solution. From an organizational perspective, security fatigue can be a serious issue and a tremendous risk. Organizations employ a variety of security measures and controls to promote secure behavior by employees (e.g., you cannot plug in external devices to corporate computers, your passwords must meet certain criteria, you are only allowed to access corporate files that are relevant to your role and/or projects, etc.). If employees become worn down or disenfranchised by all of the controls that are put in place, it is possible that they will begin ignoring security measures, circumventing controls, or, in a worst-case scenario, maliciously acting out against the organization. Practitioners must use their discretion to try to balance measures they put in place to create a secure culture within their organization without reaching the tipping point of fatigued employees who will no longer comply with policies and their accompanying controls. Such fatigue can become as serious a vulnerability as a poorly configured server or firewall. Discussion Questions

1. You have likely seen many TV shows or movies depicting hackers in futuristic looking rooms or in poorly lit basements breaking into high-security targets. Do you think these representations glorify criminal activity in a way that may entice people to engage in these activities?  Show Answer

2. One of the biggest assets that financial institutions have to secure is their reputation. If your bank or credit card company was the victim of a breach, would you have trepidation about continuing to use that financial institution? Would you switch to a competitor? What factors would inform your decision?  Show Answer

3. The article introduces security fatigue as a phenomenon that occurs when a person is so tired or overwhelmed by information security information and/or controls that they give up and actually behave insecurely (either benignly or maliciously). Do you think there would be different thresholds of security fatigue within different types of organizations?  Show Answer

4. Think about the security controls that have been put in place by your university (or employer). Do you remember the contents of the security policy that is in place? Have you ever found yourself frustrated or tired of security communications or controls that you have encountered? Explain.  Show Answer

Career Guide

Source: Colton Mouritsen, Honeywell, Inc., Senior Business Systems Analyst

· Name: Colton Mouritsen

· Company: Honeywell, Inc.

· Job Title: Senior Business Systems Analyst

· Education: Carnegie Mellon University

1. How did you get this type of job? Carnegie Mellon University (CMU) offers a plethora of unique recruiting opportunities to current students and alumni. At one of the large career fairs, I spoke to each of my target companies and was on my way out when I suddenly noticed Honeywell. Without really thinking, I walked over to the recruiter and began my pitch. She was excited about my previous work experience and current education at CMU. She informed me of a Leadership Development Program that had recently been created at Honeywell that allows individuals early in their career to gain experience across many different areas and fields in IT. I was very interested. They asked if I could interview the next day. The interview went well, and I was asked to interview with the IT/data managers with whom I would potentially work. That interview also went well, and I received an offer within a week.

2. What attracted you to this field? My path to this field is unlike most others in similar positions. I started out as an accountant and even received a Master’s in Taxation from Weber State University. During my 5-year stint in the accounting industry I learned a lot about myself and my career goals. I did not necessarily love the daily work and monotony of accounting, but I loved the side of my job where I was able to analyze data. I quickly realized that if I ever wanted to pursue a career in data analytics, I would need to further my education and develop new skills. That’s what led me to CMU. I love building models to statistically and programmatically provide insights to improve decision making. Pairing business acumen with technical skills allows real-world knowledge to combine with statistical evidence, which, in my opinion, improves decision making dramatically.

3. What does a typical workday look like for you (duties, decisions, problems)? Currently, I work on several projects that involve data science and visualization. The COVID-19 pandemic sparked a need for senior leaders at Honeywell to better understand how employees are engaging while working from home. I was given the opportunity to lead the development of the visualization. This project was requested by the chief digital technology officer and was shown to the CEO and other executives at Honeywell. My day typically involves communication with various data owners who help to validate the data that is presented in the visualization to senior leaders. I also am involved in building machine learning models for one of Honeywell’s businesses to help predict and forecast demand and sales for certain products. My teams are spread out across the world, and most of our communication is done virtually through Skype or Microsoft Teams meetings.

4. What do you like most about your job? There are complex challenges each day that I enjoy striving to solve. Learning to program has changed the way I think and problem-solve. I enjoy being presented with a problem, architecting how to programmatically solve the problem, and then actually developing the solution. I also enjoy the creativity involved in building visualizations. It can be difficult to make a visualization that not only presents accurate and easily understood data but also does so in an aesthetically pleasing way.

5. What skills would someone need to do well at your job? Being able to articulate problems and solutions is imperative in my job. Oftentimes, business leaders are not technical. It can be difficult to communicate a simple technical solution to an individual who does not understand technical terms. Also, a willingness to learn new skills on the fly has been necessary. I did not start out as a Tableau developer. In fact, I’d say my data visualization expertise was at a beginner’s level at best. The first project I was assigned was to develop Tableau dashboards. I had to learn on the fly while still producing solutions and meeting my deadlines. This can be stressful, but a desire to continue learning is extremely important in an ever-changing technical world.

6. Are education or certifications important in your field? Why? In some ways, yes. Having a great education from a top-tier institution is a great way to get into an organization. At the same time, I have noticed that once a person is able to prove their value and knowledge in building solutions and developing products, the education and certification aren’t as important. If a person can do the job well and has proven it with prior accomplishments, then certifications and education do not hold as much weight. I have seen many self-taught individuals in my career with little to no education or certifications who hold high-paying jobs and have proven to be valuable assets to their companies.

7. What advice would you give to someone who is considering working in your field? A famous cliché I live by is “there is no comfort in a growing zone, but no growing in a comfort zone.” Be willing to get out of your comfort zone and challenge yourself. Even if you fail, the experience of pushing your limits will only help to grow you as a person and in your career. Do not be scared of failure and trying new things.

8. What do you think will be hot tech jobs in 10 years? All things related to the data science industry will continue to rise. The world is already extremely connected and will only continue to be more so. Artificial intelligence is impacting nearly every industry in one way or another. Data scientists, data engineers, and machine learning engineers will be in even higher demand than they are now. Every company will need them. Using data to build models to make decisions, develop products, and automate systems/processes will no longer be considered a luxury but an essential piece to running any business, in my opinion.

Ethics Guide

Want a Loan, Who’S in your Phone Drew double-checked all of the numbers in the spreadsheet one last time in hopes that he would find an error or two, but everything looked correct. It was one of the first times in his career that he was wanting to see that he had made a mistake. No luck–everything had been entered correctly. Things at the bank were not looking good. Over the previous 3 to 6 months, there had been a pretty sharp increase in the number of loan defaults. In talking with some friends at other financial institutions, this trend did not seem to be consistent everywhere, so Drew was wondering why his bank alone would be seeing this trend.

Source: ImageFlow/Shutterstock

Just like any lender, Drew’s bank requires applicants to fill out a pile of paperwork before securing a loan. The bank looks for as much data as possible to determine creditworthiness and tries to calculate the level of risk that a borrower brings to the table. Clearly, whatever process the bank had been using to weed out risky loan applicants was not working well as it had issued a number of loans to unworthy borrowers recently–the defaults pointed to this trend. He knew he had to act quickly to change this pattern or he would be the one looking for a loan just to pay his own bills. This whole situation boiled down to getting more data–if the bank could get a better picture of who these borrowers were, it could be more accurate in predicting which ones would be more likely to default in the future. Just then, Drew remembered meeting an interesting fellow employee at the last bank-wide social event. Drew couldn’t remember his last name, but he was pretty confident his first name was Kevin. Kevin worked in the information assurance team for the bank, but aside from Kevin’s security expertise, his specialty was dealing with data. He used analytics tools to identify suspicious behaviors of employees to help identify potential insider threats. Kevin even mentioned that HR had approached him to help inform hiring processes to make sure they were onboarding the right people. Kevin hadn’t revealed much detail about the methods he used to do his work; Drew seemed to recall that Kevin may have even told him that his methods were secret. Drew didn’t see any other promising alternatives to sort this out, so he looked Kevin up in the online employee directory and started drafting a message. Drew explained the situation and told Kevin where the pending loan application files were stored in the corporate database. He finished the email by telling Kevin that anything he could do to help in terms of weeding out risky applicants would be greatly appreciated. You’ve Got Mail A few days later, Drew sat down at his desk and fired up his email client. The usual barrage of emails started popping up, but amid the flurry of emails, he noticed a response from Kevin. The first thing Drew noticed was that the message included an attachment: a spreadsheet with a list of all pending loan applications. Each name was color-coded either green or red. Aside from the list, there was nothing else included in the spreadsheet. Drew went back to the email and read the following short paragraph.

Drew, I compiled all of the pending loan applications and did an extensive analysis. In addition to the data provided by applicants, I also compiled additional data points using a variety of methods. I would prefer not going into detail, but just note that my analysis of each applicant included nontraditional factors that are now being used in other parts of the world to assess credit. I have a high degree of confidence that my recommendations are much more reliable than the traditional measures that the bank uses to identify risky applicants. Let’s leave it at that. Kevin

Drew wasn’t sure how to react to this message. Drew was aware with the fact that there are regulations in place to protect borrowers; these rules dictate the information that borrowers have to provide to secure a loan. However, he also wasn’t naive. Even as a banker, Drew realized that the real currency of today is data and that companies all over the world, including in the United States, were buying, selling, and analyzing data to better inform decision-making processes. Drew also knew that many of these companies were getting their data, and using it, in ways that may be considered questionable in terms of privacy and ethics. Drew had a hunch that Kevin’s methods included something along the lines of emerging social credit systems being developed in other countries, which, in addition to looking at traditional financial records to determine creditworthiness, now included analysis of non-financial behavior, including social media activity, relationships with family members and friends, criminal records, and so on. He couldn’t be sure about the legality of Kevin’s recommendations since Kevin didn’t give him any details about his methods. Nevertheless, Drew decided not to ask questions and opted to go ahead and use Kevin’s recommendations. He figured that if the bank kept making bad investment decisions, it wouldn’t be around long. As the only bank operation in the area, a failing bank would hurt the community as people who needed loans, and who deserved them, may not be able to get one. He started drafting emails to send out loan approval notices to the applicants highlighted in green. At this point, he was desperate to do anything to turn this loan default trend around for the best interests of his job and the community at large. Discussion Questions

1. Consider Drew’s decision to use Kevin’s loan approval recommendations.

a. Do you think the use of Kevin’s recommendations is ethical according to the categorical imperative?

b. Do you think the use of Kevin’s recommendations is ethical according to the utilitarian perspective?

2. Reports of companies’ misuse of various data sets are common, yet companies must be more efficient and effective than ever to survive. Is it justifiable for companies to “do what everyone else is doing” in order to survive, even if this behavior is questionable?

3. What would your reaction be if you learned that major financial institutions in your home country were adopting a social credit system? Would you support this initiative?

4. If you were put in charge of identifying college students who were at risk of dropping out of school, what data streams would you want to collect to most accurately identify students likely to leave the university? On the other hand, as a student, how comfortable would you be knowing that your university may be collecting numerous data streams about you to promote student retention?

Active Review

Use this Active Review to verify that you understand the ideas and concepts that answer the lesson’s study questions.

· Q3-1 How do organizations use business intelligence (BI) systems? Define business intelligence and BI system. Explain the components in Figure 3-1. Give an example, other than one in this text, of one way that an organization could use business intelligence for each of the four tasks in Figure 3-2. Name and describe the three primary activities in the BI process. Explain why master data management is important. Using Figure 3-3 as a guide, describe the major tasks for each activity. Summarize how the team at the parts distribution company used these activities to produce BI results. Explain the process shown in Figures 3-4 through 3-7.

· Q3-2 How do organizations use data warehouses and data marts to acquire data? Describe the need and functions of data warehouses and data marts. Name and describe the role of data warehouse components. Describe the differences between a data warehouse and a data lake. List and explain the problems that can exist in data used for data mining and sophisticated reporting. Use the example of a supply chain to describe the differences between a data warehouse and data mart.

· Q3-3 What are three techniques for processing BI data? Name the three types of BI analysis and describe how the goal of each differs. Name and describe five basic reporting operations in reporting analysis. Define RFM analysis and explain the actions that should be taken with customers who have the following scores: [5, 5, 5,], [1, 5, 5,], [5, 5, 3], and [5, 2, 5]. Explain OLAP and describe its unique characteristics. Explain the roles for measure and dimension in an OLAP cube. Define data mining and explain how its use typically differs from reporting applications. Describe the differences between unsupervised and supervised data mining. Explain how companies could benefit from data discovery and data visualization. Name and explain the three v’s of Big Data. Describe the general goal of MapReduce and explain, at a conceptual level, how it works. Explain the purpose of Hadoop and describe its origins. Describe the ways organizations can deploy Hadoop. Define Pig.

· Q3-4 What are the alternatives for publishing BI? Name four alternative types of servers used for publishing business intelligence. Explain the difference between static and dynamic reports; explain the term subscription. Describe why dynamic reports are difficult to create. Define knowledge management. Explain five key benefits of KM. Briefly describe one type of KM system. Summarize possible employee resistance to hyper-social knowledge sharing and name two management techniques for reducing it. Define content management system (CMS). Name two CMS application alternatives and explain the use of each. Describe five challenges organizations face for managing content.

· Q3-5 Why is artificial intelligence (AI) important? Define artificial intelligence and automation. Describe how organizations hope to use AI to increase automation. List each of the forces shown in Figure 3-23 that have driven recent advances in AI. Explain why each of these forces has been important to the current success of AI. Define deep learning. Describe how users may already be using AI applications like those shown in Figure 3-24.

· Q3-6 How will artificial intelligence and automation affect organizations? Explain how AI and automation could be used to reduce the costs and increase productivity for a fast-food restaurant. List some of the cost savings beyond wages and salaries shown in Figure 3-25 that organizations might see from adopting automated labor. Summarize some of the potential impacts of automated labor on organizational productivity as shown in Figure 3-26. Describe how the adoption of automated labor may reduce employee fraud and its potential impact on profitability. Describe two possible reactions people might have when their organizations implement widespread automation. List some jobs that humans may not want to do but that would be good for an automated worker. Describe how AI and automation will create new types of jobs to replace those that will be lost and why workers will have to adapt to these changes. List skills that can help you adapt to a shifting workplace caused by AI and automation.

· Q3-7 What is the goal of AI? Define strong AI, weak AI, and superintelligence. Describe the Turing test. Explain why the major AI research areas shown in Figure 3-30 seek to simulate all human abilities. Describe how each of these major AI research areas might be used in a self-driving car. Describe the potential effects of saying “no” to AI.

· Q3-8 How does AI work? Define machine learning, algorithm, and Naïve Bayes Classifier. Describe how machine learning uses training data to predict future outcomes. Summarize how machine learning can be used to detect spam as shown in Figure 3-31. Define natural language processing. Describe how IBM’s AI named Watson could be used by organizations to help answer user questions. Summarize how the question and answer process used by IBM’s Watson works as shown in Figure 3-32. Explain how continuous intelligence can improve existing business intelligence systems.

· Q3-9 2031? Summarize how retailers can use BI to target customers. Explain why companies in the future will know more about you than just your purchasing habits. What are some positives and negatives to analyzing this kind of data? Describe how AI might automate certain types of BI analyst jobs. Summarize the way AI could spiral out of human control. In your opinion, is this a problem? Why or why not? Describe how Kurzweil’s singularity might affect humanity.

Using Your Knowledge with eHermes From this lesson, you know the three phases of BI analysis, and you have learned common techniques for acquiring, processing, and publishing business intelligence. This knowledge will enable you to imagine innovative uses for the data that your employer generates and also to know some of the constraints of such use. At eHermes, the knowledge of this lesson will help you understand possible ways to use customer sales data to increase revenue, or possibly an AI to help optimize eHermes’ operational efficiency.

Using Your Knowledge

· 3-1. Suppose a hospital adopts a BI system paid for by a large pharmaceutical company. The system provides automated drug recommendations to doctors for each patient. According to the categorical imperative, is this ethical? Is it ethical according to utilitarianism? Do you believe it is ethical?

· 3-2. Explain in your own words how the sales analysis team in Q3-1 implemented each of the five criteria it developed. Use the data and tables shown in Q3-1 in your answer.

· 3-3. In Q3-1, the sales analysis team created a query that connected the selected parts with their past sales data (Sales History for Selected Parts). Explain why the query results do not show promise for the selling of these part designs. In light of these results, should the team look at changing its criteria? If so, how? If not, why not?

· 3-4. Given the results from the Sales History for Selected Parts query, list three actions that the company can take. Recommend one of these actions and justify your recommendation.  Show Answer

· 3-5. Describe a use for RFM analysis for Costco. Explain what you would do for customers who have the following scores: [5, 5, 5], [3, 5, 5], [5, 2, 5], [3, 3, 5], [5, 5, 3]. Is this analysis useful to Costco? Explain your answer.  Show Answer

· 3-6. Define the characteristics of Big Data. Identify and describe three student-related applications at your university that meet Big Data characteristics. Describe patterns and relationships that might be found within that data.  Show Answer

· 3-7. A robot named Vector made by Anki uses AI and neural networks to learn how to interact with people and its environment. It uses natural language processing to understand commands spoken to it and search the Internet for answers to your questions. It can also recognize specific individuals, charge itself, and navigate its way around a room. Vector can see, hear, speak, and even feel using a variety of touch sensors. Describe how AI-powered toys like Vector could change how children play with toys. How might a good experience with Vector change people’s perceptions of robots?  Show Answer

· 3-8. Thought leaders like Bill Gates, Elon Musk, and the late Stephen Hawking have expressed concerns about the potential harm that could come from AI. For example, Stephen Hawking warned that AI could evolve beyond the control of humans and cause the end of humankind. Describe some of the potential harmful effects that could come from advanced AI (superintelligence). If AI-powered robots could do 95 percent of current human jobs, what will humans do?

Collaboration Exercise

Using the collaboration IS you built in Lesson 1, collaborate with a group of students to answer the following questions. Read Case Study 3 if you have not already done so. The worldwide lockdown for COVID-19 forced many organizations to allow all of their employees to work remotely. Some workers may keep working remotely for many years. Twitter CEO Jack Dorsey announced in 2020 that all Twitter employees can work from home for as long as they want. What effect might this have on corporate culture at Twitter? Will remote employees have the same opportunities as employees who choose to go into the office? Working with your team, answer the following questions.

· 3-9. Describe how only remote collaboration might make accomplishing certain types of work more difficult. Why might face-to-face collaboration be more effective for complex, ambiguous tasks? Show Answer

· 3-10. There are many benefits of working from home, but there are downsides as well. Compare and contrast the benefits of remote work with the potential downsides. Consider impacts on the environment, organizational effectiveness, reductions in the cost of expensive office space, and so on. Show Answer

· 3-11. Consider the proverb “out of sight, out of mind.” Discuss how working from home may affect a person’s career progression. If the boss can’t see you, will your chances of a promotion be affected? If you’re not promoting yourself at work, will it be easy for contributions to be minimized or forgotten entirely? Show Answer

· 3-12. The COVID-19 lockdown of 2020 may have accelerated the eventual evolution of the modern workplace. Some workers were already working from home. Outsourcing some types of work to remote locations had been going on for years. But the lockdown forced everyone who could to work from home. Companies were forced to change their policies, workflows, and expectations. The traditional 9-to-5 workday was gone. Flexible work hours were now available to everyone. Discuss the impact of working from home on employee working conditions. Will it be easier for employees to work two or even three jobs at once? Would more effective employees make more money? Why might companies be motivated to change compensation from a “per hour” model to a “productivity” model? Show Answer

· 3-13. From a security perspective remote collaboration may cause unforeseen problems. Discuss some of the potential information security concerns that may come from having employees work from home. Consider the management and backup of data, managing corporate computing devices, and data and privacy compliance.

Case Study

We have all felt the frustration that occurs when a cell phone call becomes distorted or dropped. However, if you have ever experienced technical difficulties while videoconferencing on a work call or, even worse, during a job interview, then you know how stressful it can be. Surprisingly, unreliable and glitchy videoconferencing calls are not a thing of the past but are still quite commonplace. For some, this can be hard to believe due to how long we have been using these types of technologies. For example, Skype was a highly popular video chat platform with origins that trace all the way back to the early 2000s. However, even with almost two decades of advancements in telecommunications, Internet infrastructure, software development, and so on, you would think that a seamless videoconferencing experience would be widely available today. This has not been the case until recently.

Source: Julio Ricco/Shutterstock

Zooming to an IPO

Eric Yuan had this very thought while working for Cisco Webex–he was repeatedly told that the available solutions for videoconferencing were weak despite the fact that there were many solutions available and many companies already operating in this market space. Despite being in a high-ranking engineering position, Yuan still couldn’t convince executives to let him redesign the Cisco Webex platform to fix its unstable audio and video and lack of features.33 He was worried about a competitor developing a new and feature-rich cloud-based tool that would put the company out of business. In light of these challenges, Yuan saw an opportunity and decided to leave Cisco Webex in 2011 to start his own company. Yuan brought some engineers with him from Cisco Webex to begin working on his new platform. Finding investors to help fund the creation of this new platform was difficult as many investors felt like there was already too much competition in the videoconferencing market. After a few years of development, Yuan and his team had created a bug-resistant tool that could rapidly identify the system running the software (greatly speeding up its operations) and provide robust audio and video even with lackluster Internet connectivity. A demonstration of the ease of use and stability of the platform resulted in many interested customers–and investors–as the company progressed through each new round of fundraising. When Zoom released the first version of its new platform in 2013, it had several hundred thousand users within a month; it only took a few more months until the user count exceeded 1 million. Growth continued at breakneck pace, with user counts reaching 10 million in 2014 and 40 million in 2015. In 2017, the company received a fourth round of funding worth $100 million and was valued at $1 billion, making it a unicorn (a label for any private startup company that receives a valuation at or over $1 billion). In April 2019, Zoom had its IPO at $36 per share with a 72 percent gain on the first day of trading; after its first day as a public company, it was estimated to be worth $16 billion.

Despite Zoom’s own success, the COVID-19 pandemic in 2020 clearly contributed to a surge in its use. (As of April 2020, the company was valued at $47 billion.34) During the crisis, governments, businesses, universities, and other types of organizations all sought new digital meeting platforms so operations could continue despite people being stuck at home. However, despite the company’s successful tool and the additional interest due to COVID-19, it is not certain how long Zoom’s advantage will last. One of the growing pains Zoom experienced after its surge in users was harsh criticism about the security and privacy of its videoconferencing platform. For example, reports of Zoombombing, the act of uninvited people joining a Zoom meeting to cause disruption, became rampant in the first quarter of 2020. These meeting disruptions were also coupled with intense scrutiny about the software itself, with experts identifying a variety of weak security and privacy controls.35 Questions about Zoom started circulating. Had Zoom grown too quickly? Was the software sufficiently secure for use by massive companies and government operations? Other major tech companies detected blood in the water as they sensed an opportunity to capture market share from Zoom during this unprecedented surge in teleconferencing activity. In fact, shortly after Zoom’s rise to prominence and widespread exposure of its security and privacy concerns, Facebook, Google, Cisco, and Verizon all began tweaking and/or promoting their own videoconferencing platforms.36 How Zoom’s trajectory will change is yet to be seen. But the company’s tarnished reputation due to security flaws and a hypercompetitive market space that only appears to be heating up more may bring this company’s streak of successes to a screeching halt. Questions

· 3-14. Consider Yuan’s experience at Cisco Webex. Are you aware of any other tech companies that started out because their founder was not allowed to do something with their former employer? If not, take a few minutes to look up the histories of various tech companies to find at least one example. Show Answer

· 3-15. Zoom has been criticized for opening up its platform for use by nonpaying users (with some restrictions). Why would Zoom do this, and why would paying customers care? Show Answer

· 3-16. With the surge of Zoom users came a surge in scrutiny about the software’s privacy and security. Why do you think Zoom developers built an insecure piece of software? Show Answer

· 3-17. What is your prediction for Zoom’s long-term success? Do you think users will fall away as things normalize from COVID-19? Will competitors reveal better tools that steal market share? Explain your answer. Show Answer

· 3-18. Have you used Zoom and/or any of the competing videoconferencing platforms offered by Facebook, Google, Microsoft, and so on? How are they similar? How are they different? Which platform is your favorite? Explain your preference. Show Answer

· 3-19. Despite an already solution-rich and competitive market for videoconferencing platforms, Yuan saw an opportunity that turned into a billion-dollar company. Can you think of any markets today that may have many competitors and alternatives but products that are lacking? Show Answer

Complete the following writing exercises

· 3-20. Reflect on the differences among reporting systems, data mining systems, and Big Data systems. What are their similarities and differences? How do their costs differ? What benefits does each offer? How would an organization choose among them?

· 3-21. Install Firefox, if you do not already have it, and then install the Lightbeam add-on. Visit the sites you normally visit first thing in your day.

a. How many third-party sites are you connected to?

b. Find DoubleClick in the Lightbeam display. List the companies that DoubleClick is connected to that you did not visit.

c. Choose one of the companies in your answer to question 3-21b. Google it and describe what it does.

· 3-22. Suppose you work for an online sporting goods retailer. You’ve been hired as a business analyst with the task of increasing sales. Describe how you could use RFM to increase the sales of sporting goods. If used effectively, how could RFM affect customer satisfaction?

Lesson 9

Social Media Information Systems

Lesson Preview

Changes to social media are happening rapidly. In our experience, the best response to rapid technological change is to learn and understand underlying principles. Rather than show you Facebook or Twitter features that may change before the ink on this page is dry, let’s instead focus on principles, conceptual frameworks, and models that will be useful when you address the opportunities and risks of social media systems in the early years of your professional career. This knowledge will also help you avoid mistakes. Every day, you hear businesspeople saying, “We’re using Twitter” and “We’ve connected our Facebook page to our website.” Or they mention that they are creating ads and news releases that say, “Follow us on Twitter.” The important question is for what purpose? To be modern? To be hip? And do they have a social media strategy? Will using social media affect their bottom line? We’ll begin in Q9-1 by defining and describing the components of a social media information system, which will help you understand the commitment that organizations make when they use social media. As you’ve learned, the purpose of information systems is to help organizations achieve their strategy, and in Q9-2, we’ll consider how social media information systems facilitate organizational strategies. Next, in Q9-3, we will address how social media information systems increase social capital. Q9-4 will address how some companies earn revenue from social media, Q9-5 will look at how you can develop an effective social media strategy, and Q9-6 will look at enterprise social networks. We will then describe in Q9-7 how organizations can address security concerns related to the use of social media. We’ll wrap up in Q9-8 with an odd analogy about the change in the relationship between individuals and organizations heading into 2031.

Q9-1 What Is a Social Media Information System (SMIS)?

Social media (SM) is the use of information technology to support the sharing of content among networks of users. Social media enables people to form communities of practice, or simply communities, which are groups of people related by a common interest. A social media information system (SMIS) is an information system that supports the sharing of content among networks of users. As illustrated in Figure 9-1, social media is a convergence of many disciplines. In this course, we will focus on the MIS portion of Figure 9-1 by discussing SMIS and how they contribute to organizational strategy. If you decide to work in the SM field as a professional, you will need some knowledge of all these disciplines, except possibly computer science.

 Figure 9-1: Social Media Is a Convergence of Disciplines

Three SMIS Roles

Before discussing the components of an SMIS, we need to clarify the roles played by three organizational units:

· Social media providers

· Users

· Communities

Social Media Providers Social media providers such as Facebook, Google+, LinkedIn, Twitter, Instagram, and Pinterest provide platforms that enable the creation of social networks, or social relationships among people with common interests. The growth of SM over the past few years has been tremendous. Figure 9-2 shows the size of some well-known SM providers. In terms of the number of active users, several of these sites exceed the total population of the United States.1 The growth of SM has generated extraordinary interest from businesses, advertisers, and investors. Social media providers compete with one another for the attention of users and for the associated advertising dollars.

 Figure 9-2: Number of Social Media Active Users

Because they are information systems, SMIS have the same five components as all IS: hardware, software, data, procedures, and people. Consider each component for the roles shown in Figure 9-4.

Hardware Both users and organizations process SM sites using desktops, laptops, and mobile devices. In most cases, social media providers host the SM presence using elastic servers in the cloud. Software Users employ browsers and client applications to communicate with other users, send and receive content, and add and remove connections to communities and other users. These applications can be desktop or mobile applications for a variety of platforms, including iOS, Android, and Windows. Social media providers develop and operate their own custom, proprietary, social networking application software. As you learned in Lesson 4, supporting custom software is expensive over the long term; SM application vendors must do so because the features and functions of their applications are fundamental to their competitive strategy. They can do so because they spread the development costs over the revenue generated by millions of users. Many social networking vendors use a NoSQL database management system to process their data, though traditional relational DBMS products are used as well. Facebook began development of its own in-house DBMS (Cassandra) but later donated it to the open source community when it realized the expense and commitment of maintaining it. In addition to custom applications and databases, SM providers also invest in analytic software to understand how users interact with their site and application software. Data SM data falls into two categories: content and connections. Content data is data and responses to data that are contributed by users. You provide the source content data for your Facebook site, and your friends provide response content when they write on your wall, make comments, tag you, or otherwise publish on your site. Connection data is data about relationships. On Facebook, for example, the relationships to your friends are connection data. The fact that you’ve liked particular organizations is also connection data. Connection data differentiates SMIS from website applications. Both websites and social networking sites present user and responder content, but only social networking applications store and process connection data. SM providers store and retrieve SM data on behalf of users. They must do so in the presence of network and server failures, and they must do so rapidly. The problem is made somewhat easier, however, because SM content and connection data have a relatively simple structure. Procedures For social networking users, procedures are informal, evolving, and socially oriented. You do what your friends do. When the members of your community learn how to do something new and interesting, you copy them. SM software is designed to be easy to learn and use. Such informality makes using SMIS easy, but it also means that unintended consequences are common. The most troubling examples concern user privacy. Many people have learned not to post pictures of themselves in front of their house numbers on the same publicly accessible site on which they’re describing their new high-definition television. Many others, alas, have not. For organizations, social networking procedures are more formalized and aligned with the organization’s strategy. Organizations develop procedures for creating content, managing user responses, removing obsolete or objectionable content, and extracting value from content. For example, setting up an SMIS to gather data on product problems is a wasted expense unless procedures exist to extract knowledge from that social networking data. Organizations also need to develop procedures to manage SM risk, as described in Q9-7. Procedures for operating and maintaining the SM application are beyond the scope of this text. People Users of social media do what they want to do depending on their goals and their personalities. They behave in certain ways and observe the consequences. They may or may not change their behavior. By the way, note that SM users aren’t necessarily rational, at least not in purely monetary ways. See, for example, the study by Vernon Smith in which people walked away from free money because they thought someone else was getting more! 7  Organizations cannot be so casual. Anyone who uses his or her position in a company to speak for an organization needs to be trained on both SMIS user procedures and the organization’s social networking policy. We will discuss such procedures and policies in Q9-7. Social media is creating new job titles, new responsibilities, and the need for new types of training. For example, what makes a good tweeter? What makes an effective wall writer? What type of people should be hired for such jobs? What education should they have? How does one evaluate candidates for such positions? How do you find these types of people? All of these questions are being asked and answered today.

Q9-2 How Do SMIS Advance Organizational Strategy?

In Lesson 2, Figure 2-1, you learned the relationship of information systems to organizational strategy. In brief, strategy determines value chains, which determine business processes, which determine information systems. Insofar as value chains determine structured business processes, such as those discussed in Lesson 8, this chain is straightforward. However, social media is by its very nature dynamic; its flow cannot be designed or diagrammed, and if it were, no sooner would the diagram be finished than the SM process would have changed. Therefore, we need to back up a step and consider how value chains determine dynamic processes and thus set SMIS requirements. As you will see, social media fundamentally changes the balance of power among users, their communities, and organizations. Figure 9-5 summarizes how social media contributes to the five primary value chain activities and to the human resources support activity. Consider each row of this table.

In the past, organizations controlled their relationships with customers using structured processes and related information systems. In fact, the primary purpose of traditional CRM was to manage customer touches. Traditional CRM ensured that the organization spoke to customers with one voice and that it controlled the messages, the offers, and even the support that customers received based on the value of a particular customer. In 1990, if you wanted to know something about an IBM product, you’d contact its local sales office; that office would classify you as a prospect and use that classification to control the literature, the documentation, and your access to IBM personnel. Social CRM is a dynamic, SM-based CRM process. The relationships between organizations and customers emerge in a dynamic process as both parties create and process content. In addition to the traditional forms of promotion, employees in the organization create wikis, blogs, discussion lists, frequently asked questions, sites for user reviews and commentary, and other dynamic content. Customers search this content, contribute reviews and commentary, ask more questions, create user groups, and so forth. With social CRM, each customer crafts his or her own relationship with the company. Social CRM flies in the face of the structured and controlled processes of traditional CRM. Because relationships emerge from joint activity, customers have as much control as companies. This characteristic is anathema to traditional sales managers who want structured processes for controlling what the customer reads, sees, and hears about the company and its products. Further, traditional CRM is centered on lifetime value; customers who are likely to generate the most business get the most attention and have the most effect on the organization. However, with social CRM, the customer who spends 10 cents but who is an effective reviewer, commentator, or blogger can have more influence than the quiet customer who purchases $10M a year. Such imbalance is incomprehensible to traditional sales managers. However, traditional sales managers are happy to have loyal customers sell their products using peer-to-peer recommendations. A quick look at products and their reviews on Amazon will show how frequently customers are willing to write long, thoughtful reviews of products they like or do not like. Amazon and other online retailers also allow readers to rate the helpfulness of reviews. In that way, substandard reviews are revealed for the wary. Today, many organizations are struggling to make the transition from controlled, structured, traditional CRM processes to wide-open, adaptive, dynamic social CRM processes; this struggle represents a significant job opportunity for those interested in IS, sales, and social media.

Social Media and Customer Service

Product users are amazingly willing to help each other solve problems. Even more, they will do so without pay; in fact, payment can warp and ruin the support experience as customers fight with one another. SAP, for example, learned that it was better to reward its SAP Developer Network with donations on their behalf to charitable organizations than to give them personal rewards. Not surprisingly, organizations whose business strategy involves selling to or through developer networks have been the earliest and most successful at SM-based customer support. In addition to SAP, Microsoft has long sold through its network of partners. Its MVP (Most Valuable Professional) program is a classic example of giving praise and glory in exchange for customer-provided customer assistance (http://mvp.support.microsoft.com). Of course, the developers in these networks have a business incentive to participate because that activity helps them sell services to the communities in which they participate. However, users with no financial incentive are also willing to help others. For instance, Amazon supports a program called Vine by which customers can be selected to give prerelease and new product reviews to the buyer community.8 You’ll need your psychology course to explain what drives people to strive for such recognition. MIS just provides the platform! The primary risk of peer-to-peer support is loss of control. Businesses may not be able to control peer-to-peer content. Negative comments about cherished products and recommendations for competitor’s products are a real possibility. We address these risks in Q9-7.

Companies whose profitability depends on the efficiency of their supply chain have long used information systems to improve both the effectiveness and efficiency of structured supply chain processes. Because supply chains are tightly integrated into structured manufacturing processes, there is less tolerance for the unpredictability of dynamic, adaptive processes. Solving problems is an exception; social media can be used to provide numerous solution ideas and rapid evaluation of them. The Japanese earthquake in the spring of 2011 created havoc in the automotive supply chain when major Japanese manufacturers lacked power and, in some cases, facilities to operate. Social media was used to dispense news, allay fears of radioactive products, and address ever-changing needs and problems. SM communities may provide better and faster problem solutions to complex supply chain problems. Social media is designed to foster content creation and feedback among networks of users, and that characteristic facilitates the iteration and feedback needed for problem solving, as described in Lesson 7. Loss of privacy is, however, a significant risk. Problem solving requires the open discussion of problem definitions, causes, and solution constraints. Because suppliers and shippers work with many companies, supply chain problem solving via social media may be problem solving in front of your competitors.

Social Media and Manufacturing and Operations

Operations and manufacturing activities are dominated by structured processes. The flexibility and adaptive nature of social media would result in chaos if applied to the manufacturing line or to the warehouse. However, social media does play a role in designing products, developing supplier relationships, and improving operational efficiencies. Crowdsourcing is the dynamic social media process of employing users to participate in product design or product redesign. eBay often solicits customers to provide feedback on their eBay experience. As its site says, “There’s no better group of advisors than our customers.” User-guided design has been used to create video games, shoes, and many other products. Social media has been widely used in business-to-consumer (B2C) relationships to market products to end users. Now manufacturers are starting to use social media to become industry leaders, promote brand awareness, and generate new business-to-business (B2B) leads to retailers. Manufacturers can use social media by starting a blog that discusses the latest industry-related news, posts interviews with experts, and comments on new product innovations. They can also create a YouTube channel and post videos of product reviews and testing and factory walk-throughs. Facebook and Twitter accounts are useful to promote positive consumer stories, announce new products, and follow competitors. Retailers view manufacturers who engage in such SM efforts as industry leaders. Operations can use social media to improve communication channels within the organization and externally with consumers. For example, an enterprise social networking service like Yammer can be used to provide managers with real-time feedback about how to resolve internal operational inefficiencies. Externally, a retailer could monitor its corporate Twitter account and respond to product shortages or spikes in demand for new products around holidays.

The last row in Figure 9-5 concerns the use of social media in human resources. As previously mentioned, social media is used for finding employee prospects, for recruiting candidates, and—in some organizations—for candidate evaluation. Organizations use social media sites like LinkedIn to hire the best people more quickly and at a lower cost. For about $900 a month, recruiters can search through 660 million LinkedIn members to find the perfect candidate.9 That $900 a month may sound like a lot to you, but to corporate customers, it’s peanuts. The cost of hiring just one new employee runs around $5,000.10 If an independent recruiting company is involved, that cost can be as high as 10 percent of the new employee’s salary. LinkedIn also gives employers access to passive candidates who might not be looking for a job but are a perfect fit for a particular position. Once the employee is hired, the employer can leverage that new employee’s social network to hire more candidates just like him or her. Jobvite, a social recruiting company, reports that 96 percent of recruiters it surveyed used social media in their recruiting process. Furthermore, 60 percent of recruiters reported that they view shared details about volunteer or social engagement work on social media sites positively. However, posts with spelling mistakes or poor grammar were viewed negatively by 43 percent of respondents. They also viewed posts about alcohol consumption (35 percent) and marijuana use (61 percent) negatively.11 Social media is also used for employee communications, using internal personnel sites such as MySite and MyProfile in SharePoint or other similar enterprise systems. SharePoint provides a place for employees to post their expertise in the form of “Ask me about” questions. When employees are looking for an internal expert, they can search SharePoint for people who have posted the desired expertise. SharePoint 2019 greatly extends support for social media beyond that in earlier SharePoint versions. The risks of social media in human resources concern the possibility of error when using sites such as Facebook to form conclusions about employees. A second risk is that the SM site becomes too defensive or is obviously promulgating an unpopular management message. Study Figure 9-5 to understand the general framework by which organizations can accomplish their strategy via a dynamic process supported by SMIS. We will now turn to an economic perspective on the value and use of SMIS.

Q9-3 How Do SMIS Increase Social Capital?

Business literature defines three types of capital. Karl Marx defined capital as the investment of resources for future profit. This traditional definition refers to investments into resources such as factories, machines, manufacturing equipment, and the like. Human capital is the investment in human knowledge and skills for future profit. By taking this class, you are investing in your own human capital. You are investing your money and time to obtain knowledge that you hope will differentiate you from other workers and ultimately give you a wage premium in the workforce. According to Nan Lin, social capital is the investment in social relations with the expectation of returns in the marketplace.12 You can see social capital at work in your personal life. You strengthen your social relationships when you help someone get a job, set a friend up on a date, or introduce a friend to someone famous. You weaken the strength of your social relationships by continually freeloading, declining requests for help, and failing to spend time with friends. In your professional life, you are investing in your social capital when you attend a business function for the purpose of meeting people and reinforcing relationships. Similarly, you can use social media to increase your social capital by recommending or endorsing someone on LinkedIn, liking a picture on Facebook, retweeting a tweet, or commenting on an Instagram picture.

According to Lin, social capital adds value in four ways:

· Information

· Influence

· Social credentials

· Personal reinforcement

First, relationships in social networks can provide information about opportunities, alternatives, problems, and other factors important to business professionals. On a personal level, this could come in the form of a friend telling you about a new job posting or the best teacher to take for Business Law. As a business professional, this could be a friend introducing you to a potential new supplier or letting you know about the opening of a new sales territory. Second, relationships provide an opportunity to influence decision makers at your employer, or in other organizations, who are critical to your success. For example, playing golf every Saturday with the CEO of the company you work for could increase your chances of being promoted. Such influence cuts across formal organizational structures, such as reporting relationships. Third, being linked to a network of highly regarded contacts is a form of social credential. You can bask in the glory of those with whom you are related. Others will be more inclined to work with you if they believe critical personnel are standing with you and may provide resources to support you. Finally, being linked into social networks reinforces a professional’s identity, image, and position in an organization or industry. It reinforces the way you define yourself to the world (and to yourself). For example, being friends with bankers, financial planners, and investors may reinforce your identity as a financial professional. As mentioned, a social network is a network of social relationships among individuals with a common interest. Each social network differs in value. The social network you maintain with your high school friends probably has less value than the network you have with your business associates, but not necessarily so. According to Henk Flap, the value of social capital is determined by the number of relationships in a social network, by the strength of those relationships, and by the resources controlled by those related.13 If your high school friends happened to have been Mark Zuckerberg or Cameron and Tyler Winklevoss and if you maintain strong relations with them via your high school network, then the value of that social network far exceeds any you’ll have at work. For most of us, however, the network of our current professional contacts provides the most social capital. So, when you use social networking professionally, consider these three factors. You gain social capital by (1) adding more friends, (2) strengthening the relationships you have with existing friends, and (3) strengthening relationships with people who control resources that are important to you. Such calculations may seem cold, impersonal, and possibly even phony. When applied to the recreational use of social networking, they may be. But when you use social networking for professional purposes, keep them in mind. As a business professional, it’s important to understand what social capital is, why it’s valuable, and how you can benefit from it.

Organizations have social capital just as humans do. Historically, organizations created social capital via salespeople, customer support, and public relations. Endorsements by high-profile people are a traditional way of increasing social capital, but there are tigers in those woods. Today, progressive organizations maintain a presence on Facebook, LinkedIn, Twitter, and possibly other sites. They include links to their social networking presence on their websites and make it easy for customers and interested parties to leave comments. To understand how social networks add value to businesses, consider each of the elements of social capital: number of relationships, strength of relationships, and resources controlled by “friends.”

In a traditional business relationship, a client (you) has some experience with a business, such as a restaurant or resort. Traditionally, you may express your opinions about that experience by word of mouth to your social network. If you are an influencer in your social network, your opinion may force a change in others’ behavior and beliefs. However, such communication is unreliable and brief: You are more likely to say something to your friends if the experience was particularly good or bad, but even then, you are likely only to say something to those friends whom you encounter while the experience is still recent. And once you have said something, that’s it; your words don’t live on for days or weeks. However, what if you could use SM to communicate your experience using text, pictures, and video instantly to everyone in your social network? For example, suppose a wedding photographer uses social media to promote her business by asking a recent client (user 1) to “like” her Facebook page and the wedding photos posted there (Figure 9-6). She also tags people in the client’s pictures on Facebook. She may even ask the client to tweet about her experience.

Figure 9-6: Growing Social Networks All of the people in the client’s social network (users 4–6) see the likes, tags, and tweets. If user 6 likes the pictures, they might be seen by users 10–12. It’s possible that one of those users is looking for a wedding photographer. Using social media, the photographer has thus grown her social network to reach potential clients who she wouldn’t have otherwise had access to. She also used SM to grow the number of relationships she has with clients. Depending on the number, strength, and value of those relationships, her social capital within those networks could substantially increase. Such relationship sales have been going on by word of mouth for centuries; the difference here is that SMIS allow such relationships to scale to levels not possible in the past. In fact, the photographer in our example might even consider paying the client for the opportunity to take the wedding pictures if the client were a famous celebrity with hundreds of thousands of followers. In this way, social media may allow users to convert social capital into financial capital. Some famous celebrities get paid more than $1,000,000 for an Instagram post14 and $30,000 for a single tweet!15

To an organization, the strength of a relationship is the likelihood that the other entity (person or other organization) in the relationship will do something that benefits the organization. An organization may have a strong relationship with you if you write positive reviews about it, post pictures of you using the organization’s products or services, tweet about upcoming product releases, and so on. In the previous example, the photographer asked a client to like her Facebook page and wedding photos. To the photographer, the number of friends the client has in her social network is important, but equally important is the strength of the relationships. Will the client’s friends like the photographer’s page and photos? Will they retweet the client’s success story? If none of the client’s friends like the photographer’s page and photos, then the strength of the relationships is weak. If all of the client’s friends like the photographer’s page and photos, then the strength of the relationships in the client’s social network is strong. In his autobiography, Benjamin Franklin provided a key insight. He said that if you want to strengthen your relationship with someone in power, ask him to do you a favor. Before Franklin invented the public library, he would ask powerful strangers to lend him their expensive books. In that same sense, organizations have learned that they can strengthen their relationships with you by asking you to do them a favor. When you provide that favor, it strengthens your relationship with the organization. Traditional capital depreciates. Machines wear out, factories get old, technology and computers become obsolete, and so forth. Does social capital also depreciate? Do relationships wear out from use? So far, the answer seems to be both yes and no. Clearly, there are only so many favors you can ask of someone in power. And there are only so many times a company can ask you to review a product, post pictures, or provide connections to your friends. At some point, the relationship deteriorates due to overuse. So, yes, social capital can be spent. However, frequent interactions strengthen relationships and hence increase social capital. The more you interact with a company, the stronger your commitment and allegiance. But continued frequent interactions occur only when both parties see value in continuing the relationship. Thus, at some point, the organization must provide you an incentive to continue to do it a favor. So, social capital can be spent, but it can also be earned by adding something of value to the interaction. If an organization can induce those in its relationships to provide more influence, information, social credentials, or personal reinforcement, it has strengthened those relationships. And continuing a successful relationship over time substantially increases relationship strength.

The third measure of the value of social capital is the value of the resources controlled by those in the relationships. An organization’s social capital is thus partly a function of the social capital of those to whom it relates. The most visible measure is the number of relationships. Someone with 1,000 loyal Twitter followers is usually more valuable than someone with 10. But the calculation is subtler than that; for example, if those 1,000 followers are college students and if the organization’s product is adult diapers, then the value of the relationship to the followers is low. A relationship with 10 Twitter followers who are in retirement homes would be more valuable.

Negative customer reviews on social media sites can harm businesses financially. Manipulating reviews can be a questionable practice. Read more about it in the Ethics Guide.

To illustrate this point, Figure 9-7 shows rankings of YouTube channels by annual earnings, number of subscribers, and annual views.16 Note that the channel with the highest annual earnings (Ryan ToysReview) does not have the most subscribers (T-Series) or the most annual views (Music). The resources (i.e., money) controlled by the viewers of the Ryan ToysReview channel are highly valued by paying advertisers even though Ryan ToysReview only ranks fourth in the number of annual views and 103rd in number of subscribers. 

There is no formula for computing social capital, but the three factors would seem to be more multiplicative than additive. Or, stated in other terms, the value of social capital is more in the form of Social Capital = Number of Relationships * Relationship Strength * Entity Resources than in the form of Social Capital = Number of Relationships + Relationship Strength + Entity Resources Again, do not take these equations literally; take them in the sense of the multiplicative interaction of the three factors. This multiplicative nature of social capital means that a huge network of relationships with people who have few resources may be of less value than a smaller network of relationships with people who have substantial resources. Furthermore, those resources must be relevant to the organization. Students with pocket change are relevant to Pizza Hut; they are irrelevant to a BMW dealership. This discussion brings us to the brink of social networking practice. Most organizations today ignore the value of entity assets and simply try to connect to more people with stronger relationships. This area is ripe for innovation. Data aggregators such as ChoicePoint and Acxiom maintain detailed data about people worldwide. It would seem that such data could be used by information systems to calculate the potential value of a relationship to a particular individual. This possibility would enable organizations to better understand the value of their social networks as well as guide their behavior with regard to particular individuals. Stay tuned; many possibilities exist, and some ideas—maybe yours—will be very successful.

Knowledge Check

Q9-4 How Do (Some) Companies Earn Revenue from Social Media?

Having a large social network with strong relationships may not be enough. Facebook, for example, has more than 2.5 billion active users that share over 4.75 billion pieces of content each day.17 YouTube has more than 2 billion active users that watch more than 1 billion hours of video each day.18 Both companies have extremely large numbers of active users. The only problem is that they give it away for free. Billions of anything multiplied by zero is zero. Do all those users really matter if Facebook and YouTube can’t make a single penny off of them? As a business student, you know that nothing is free. Processing time, data communication, and data storage may be cheap, but they still cost something. Who pays for the hardware? Social media companies like Facebook, Twitter, and LinkedIn also need to pay people to develop, implement, and manage the SMIS. And where does Web content come from? Fortune pays authors for the content that it offers for free. Who is paying those authors? And from what revenue?

Social media has evolved in such a way that users expect to use SM applications without paying for them. SM companies want to build up a large network of users quickly, but they have to offer a free product in order to attract users. The dilemma then becomes how do they monetize, or make money from, their application, service, or content. The answer is by making users the product. That may sound strange at first. You don’t want to think of yourself as a product. But try to look at it from the company’s point of view. When a company runs an advertisement, it’s essentially being paid to put the ad in front of its users. In a way, it’s renting your eyeballs to an advertiser for a short period of time. Google is paid to target users with ads by using their search terms, sites they visit, and “scans” of their emails to place targeted ads in front of them. In essence, then, users are the product being sold to advertisers. As the old saying says, “If you’re not paying, you’re the product.”

The two most common ways SM companies generate revenue are advertising and charging for premium services. On Facebook, for example, creating a company page is free, but Facebook charges a fee to advertise to communities that “like” that page. Advertising Most SM companies earn revenue through advertising. Facebook made 98 percent of its 2020 first quarter earnings (.8B) from advertising.19 About 84 percent of Twitter’s 2M first quarter earnings came from advertising as well.20 Advertising on SM can come in the form of paid search, display or banner ads, mobile ads, classifieds, or digital video ads. Google led the way in making digital advertising revenue with search, followed by Gmail and then YouTube. Today, it doesn’t seem like any great insight to realize that if someone is searching for information about an Audi A5 Cabriolet, then that person may be interested in ads from local Audi dealers and BMW and Mercedes dealers as well. Or if someone is watching a soccer game on YouTube, maybe he or she likes soccer. While not mind-boggling to imagine, Google was the first to turn this notion into substantial revenue streams. Other tech companies followed. Advertisers like digital ads because, unlike traditional media such as newspapers, users can respond directly to Web ads by clicking on them. Run an ad in the print version of The Wall Street Journal, and you have no idea who responds to that ad or how strongly. But place an ad for that same product in the newspaper’s online version, and you’ll soon know the percentage of viewers who clicked that ad and what action they took next. This knowledge led to the pay-per-click revenue model, in which advertisers display ads to potential customers for free and pay only when the customer clicks. Another way to grow ad revenue is to increase site value with user contributions. The term use increases value means the more people use a site, the more value it has, and the more people will visit. Furthermore, the more value a site has, the more existing users will return. This phenomenon led to user comments and reviews, blogging, and, within a few years, social media. If you can get people to connect their community of practice to a site, you will get more users, they will add more value, existing users will return more frequently, and, all things considered, the more ad clicks there will be. Freemium The freemium revenue model offers users a basic service for free and then charges a premium for upgrades or advanced features. LinkedIn earns part of its revenue by selling upgrades to its standard software as a service (SaaS) product. As of May 2020, regular users access LinkedIn for free; individual upgrades range from to a month and offer advanced search capabilities, greater visibility of user profiles, and more direct email messages to LinkedIn users outside one’s network. Businesses that want to use LinkedIn for recruiting can purchase a Recruiter Corporate account for 0 to 00 a month. LinkedIn’s revenue consists of about 17 percent from premium subscriptions, 65 percent from online recruitment, and 18 percent from advertising.21 By diversifying its revenue streams, LinkedIn has reduced its dependence on fluctuating ad revenue and lessened the negative impact of ad-blocking software. A recent report by Blockthrough found that 236 million desktop users and 527 mobile users are actively using ad-blocking software to filter out advertising content. They rarely, if ever, see Internet ads.22 It also reported that 69 percent of global ad blocking is done on mobile devices. SM companies that rely solely on ad revenue may see their share prices plummet if the use of ad-blocking software becomes widespread. Other ways of generating revenue on SM sites include the sale of apps and virtual goods, affiliate commissions, and donations. During the month of April 2018, the free-to-play game Fortnite generated 6 million from the sale of virtual goods. Wikipedia took in about .7M in donations during 2018.23 Interestingly, some SM companies generate revenue but don’t make any profit. Pinterest, for example, made .14 billion in revenue in 2019. Unfortunately, Pinterest’s expenses were .53 billion, resulting in a .36 billion net loss. Some SM companies just focus on building a large network of users now and figuring out how to make money later. Social media is the ultimate expression of use increasing value. The more communities of practice there are, the more people, and the more incentive people will have to come back again and again. So, social media would seem to be the next great revenue generator, except, possibly, for the movement from PCs to mobile devices.

The ad click revenue model successfully emerged on PC devices where there is plenty of space for lots of ads. However, as users move from PCs to mobile devices, particularly small-screen smartphones, there is much less ad space. Does this mean a reduction in ad revenue? On the surface, yes. According to eMarketer, mobile ad spending will increase more than 20 percent in 2020 to $105B accounting for 70 percent of total digital ad spending.24 By 2023, as shown in Figure 9-8, mobile ad spending should reach $155B and account for 77 percent of total digital ad spending. However, growth in the number of mobile devices far exceeds PC growth.

 Figure 9-8: Mobile Ad Spending

Q9-5 How Do Organizations Develop an Effective SMIS?

At this point in your reading, you know what SMIS are, why they are important, and how they generate revenue. Now you need to know how to develop an effective SMIS that is strategically aligned with your organization’s goals. In Q9-2, you saw that SM can be used to benefit an organization, but how do you get to that point? We’re not talking about a recipe for turning your organization into the next Facebook. Rather, the steps shown in Figure 9-9 walk you through the process of developing a practical plan to effectively use existing social media platforms.

 Figure 9-9: Social Media Plan Development

Social media has been criticized because of potentially deceptive tactics used by influencers and companies. Read more about this in the Security Guide.

Many companies are still unsure how to use SM. They want to use it, but they’re unsure how to facilitate their existing competitive strategy. Think back to Porter’s model for competitive strategies from Lesson 2 (Figure 2-5). Organizations can focus their strategies on being the cost leader or on differentiating their products from the competition. Organizations can then employ the chosen strategy across an entire industry or focus on a particular segment within that industry. Depending on an organization’s strategy, it will use different SM platforms in different ways. Again, the key is premeditated alignment of the SMIS with the organization’s strategy. Organizations know SM is popular and could be strategically beneficial. They hear about it constantly in the news. It’s not entirely their fault if they want to jump on board. Social media is a relatively new development with a dizzying array of companies, platforms, and services. It’s constantly changing, too. It’s important to understand the development process presented in Figure 9-9 because you may be the “social media expert” at your future job. You may be called in to help develop the organization’s SMIS. In order to be successful, take a few minutes to consider the steps in the process.

It may sound cliché, but the first step in developing an SMIS is to clearly define what the organization wants to achieve with SM. As previously mentioned, your goals must be clear, deliberate, and aligned with the organization’s competitive strategy. Without clearly defined goals, you won’t know whether your SM effort was successful. As you learned in Lesson 2, the goals for each organization are different. For organizations that choose a differentiation strategy, SM goals could include better employee recruiting, quicker product development, becoming an industry product leader, or increasing customer loyalty. In general, most organizations include increased brand awareness, conversion rates, website traffic, or user engagement as goals. Figure 9-10 gives you examples of how these might manifest themselves in social media.

Step 2: Identify Success Metrics

After you know what you want to accomplish using SM, you need to identify metrics that will indicate when you’ve achieved your goals. These are referred to as success metrics or key performance indicators (KPI). Metrics are simply measurements used to track performance. Every organization has different metrics for success. For example, a law firm may measure billable hours, a hospital may measure patients seen or procedures performed, and a manufacturer may look at units produced or operational efficiency. The hard part in identifying success metrics is identifying the right ones. The right metrics help you make better decisions; the wrong metrics are meaningless and don’t positively affect your decision making. For example, measuring the number of registered users on your site may be interesting but not really meaningful. What really matters is the number of active users on your site each month. For example, over 1.3 billion Twitter accounts have been created, but only 700 million accounts are active per year. Even more interesting, Twitter has 330 million monthly active users, and 145 million daily users.30 Which metric is the most useful? The metrics that describe the number of active users most accurately are probably the monthly active users, and daily active users. Total accounts created wouldn’t be a good success metric because it’s not very meaningful. Metrics that don’t improve your decision making are commonly referred to as vanity metrics. Figure 9-11 shows examples of possible success metrics for the goals mentioned in Figure 9-10. Remember, in some circumstances you want to maximize the metric, while in others you want to minimize the metric. It’s similar to sports in that respect: Sometimes you want a high score (basketball), and other times you want a low score (golf). It just depends on what you’re measuring. Whereas you may want to maximize a metric like conversion rate,31 or the percentage of people who achieve a certain result, you will probably want to minimize other metrics like bounce rate, or the percentage of people who visit your website and then immediately leave.

The next step in creating an effective SMIS is to clearly identify your target audience. Chances are it’s not going to be everyone. For example, if you’re Caterpillar Inc. trying to use social media to sell more D11 dozers, your target audience probably won’t include many teenagers. Organizations go to great lengths to identify their target audience because it helps them focus their marketing efforts. Once you’ve identified your target audience, you need to find out which SM platforms they use. Certain social media platforms attract certain audiences. For example, more than 71 percent of Pinterest users are women,32 65 percent of Instagram users are between 18 and 34 years old,33 and 45 percent of LinkedIn users earn over $75,000 per year.34 Your target audience will influence which SM platforms you use.

After pinpointing your target audience, you’ll need to define the value you’ll provide your audience. Why should these users listen to you, go to your website, like your posts, or tweet about your products? Are you providing news, entertainment, education, employee recruiting, or information? In essence, you need to define what you are going to give your audience in exchange for making a connection with you.

See what a typical workday would look like for someone who manages social media in the Career Guide.

Shopping is a good metaphor to explain how you can do this. When you go shopping, you see something of value and you exchange your financial capital (money) with the business for the item you value. The same is true of social media. Your audience members are constantly browsing for things of value, and they have social capital to spend. They may eventually spend financial capital at your website, but it’s the social capital that is most important. You need to define what you’re going to offer users in exchange for their social capital. Take LinkedIn as an example. It helps users find jobs, build a professional network, join special interest groups, get introduced to prospective clients, and reconnect with past colleagues. From an organizational perspective, LinkedIn allows recruiters to quickly identify and contact potential hires from a large pool of candidates. This lowers hiring costs and improves the quality of new hires. If you’re unsure how your organization could add value, start by performing a competitive analysis to identify the strengths and weaknesses in your competitors’ use of social media. Look at what they’re doing right and what they’re doing wrong.

The true value of social media can be achieved only when organizations use social media to interact with customers, employees, and partners in a more personal, humane, relationship-oriented way. According to recent studies, younger users are more skeptical of organizational messages and may no longer listen to them. A CivicScience study found that 58 percent of younger consumers ages 18 through 29 were more influenced by social media chatter than either TV ads or Internet ads.35 Interestingly, the study also found that only 29 percent of consumers over age 55 thought social media chatter was more influential than TV or Internet advertising. Such skepticism by younger consumers is understandable. They grew up with more sources of information and feel comfortable using social media. Skepticism of organizational messages gives a competitive advantage to organizations that can make personal connections with users via social media. Today, people want informed, useful interactions that help them solve particular problems and satisfy unique needs. They increasingly ignore prepackaged organizational messages that tout product benefits. This requires you to engage audience members, ask them questions, and respond to their posts. It also means you must avoid hard-selling products, overwhelming audience members with content, and contacting them too often. The sales force in Apple stores is an excellent example of how to make personal connections. Team members have been trained to act as customer problem-solving consultants and not as sellers of products. An organization’s use of social media needs to mirror this behavior; otherwise, social media is nothing more than another channel for classic advertising.

Finally, when creating a social media strategy, you need to gather the right amount of data necessary to make the most informed decision you can. You can use online analytical tools like Google Analytics, Facebook Page Insights, Clicky, or KISSmetrics to measure the success metrics you defined earlier. These tools will show you statistical information such as which tweets get the most attention, which posts generate the most traffic, and which SM platform generates the most referrals. Then you can refine your use of social media based on the performance of your success metrics. Be sure to rely on analysis of hard data, not anecdotes from friends. Also, remember that the SM landscape is changing rapidly, and today’s winners could be tomorrow’s losers. MySpace, for example, was the top SM site in late 2007 valued at $65B, but then succumbed to Facebook’s success and was sold for $35M in 2011.36 Users may shift away from current SM giants like Facebook toward a group of more customized applications like Instagram, Twitter, Snapchat, and WhatsApp.37 Allow your use of social media to be flexible enough to change with the times. Senior managers need to see regular progress reports about how SM is affecting the organization. They also need to be educated about changes in social media landscape. Watch for SM success stories and communicate them with upper management.

Knowledge Check

Q9-6 What Is an Enterprise Social Network (ESN)?

An enterprise social network (ESN) is a software platform that uses social media to facilitate cooperative work of people within an organization. Instead of using outward-facing SM platforms like Facebook and Twitter, it uses specialized enterprise social software designed to be used inside the organization. These applications may incorporate the same functionality used by traditional social media, including blogs, microblogs, status updates, image and video sharing, personal sites, and wikis. The primary goal of enterprise social networks is to improve communication, collaboration, knowledge sharing, problem solving, and decision making.

Enterprise 2.0

In 2006, Andrew McAfee wrote an article about how dynamic user-generated content systems, then termed Web 2.0, could be used in an enterprise setting. He described Enterprise 2.0 as the use of emergent social software platforms within companies.38 In other words, the term Enterprise 2.0 refers to the use of enterprise social networks. McAfee defined six characteristics that he refers to with the acronym SLATES (see Figure 9–12).39 First, workers want to be able to search for content inside the organization just like they do on the Web. Most workers find that searching is more effective than navigating content structures such as lists and tables of content. Second, workers want to access organizational content via links, just as they do on the Web. They also want to author organizational content using blogs, wikis, discussion groups, published presentations, and so on.

According to McAfee, a fourth characteristic of ESNs is that their content is tagged, just like content on the Web, and these tags are organized into structures, as is done on the Web at sites like Delicious (www.delicious.com). These structures organize tags as a taxonomy does, but unlike taxonomies, they are not preplanned; they emerge organically. In other words, ESNs employ a folksonomy, or a content structure that emerges from the processing of many user tags. Fifth, workers want applications that enable them to rate tagged content and to use the tags to predict content that will be of interest to them (as with Pandora), a process McAfee refers to as extensions. Finally, workers want relevant content pushed to them; or, in McAfee’s terminology, they want to be signaled when something that interests them happens in organizational content. The potential problem with ESNs is the quality of their dynamic process. Because the benefits of an ESN result from emergence, there is no way to control for either effectiveness or efficiency. It’s a messy process about which little can be predicted.

Prior to 1980, communication in the United States was restricted to a few communication channels, or means of delivering messages. There were three major national TV networks and no more than a half-dozen major national newspapers. Consumers got their news twice a day: from the morning paper and the evening news. A small number of people decided which stories were told. You got what you were given with few alternatives. Communication within organizations was similarly restricted. Employees could communicate with their immediate supervisor and coworkers in their vicinity. It was difficult for employees of large corporations to get private meetings with the CEO or to communicate with their counterparts in other countries. If an employee had a good idea, it was passed up through his or her boss to senior management. As a result, it was common for bosses to claim subordinates’ ideas as their own. In recent decades, the Internet, websites, social networking, email, cable TV, and smartphones have radically altered existing communication channels. At the societal level, you can now get your news instantly from hundreds of different sources. Traditional news organizations have struggled to adapt to changes in traditional communication channels. Communication channels within corporations have changed in equally dramatic ways. Using ESNs, employees can now bypass managers and post ideas directly for the CEO to read. They can also quickly identify internal subject matter experts to solve unforeseen problems. In addition, ESNs also enable collaboration with teams dispersed across the globe. To better understand the potential impacts of ESNs, let’s consider an example. In 2017, Cummins, a large U.S.-based manufacturer, implemented Yammer (a Microsoft subsidiary) to help its 59,000 employees communicate more effectively. Today, more than 17,000 employees use Yammer to communicate across their global enterprise social network.40 Cummins employees use the translation feature in Yammer to automatically communicate in 19 different languages. In another example, the CIO of Red Robin, Chris Laping, rolled out Yammer to Red Robin’s 26,000 employees across 450 restaurants in an effort to give line employees a voice. Laping offered a $1,000 employee bonus for the best cost-saving idea. The winning idea was reusable kids’ cups that saved hundreds of thousands of dollars. Laping attributes the cost savings to the ESN, stating, “I’m convinced that idea would never have surfaced if we didn’t have a social network.”41

The use of ESNs in organizations is new, and organizations are still learning how to use ESNs successfully (creating fascinating job opportunities for you, by the way). Before deploying an ESN, organizations should develop a strategic plan for using SM internally via the same process they used for their external social media use. Once a strategic plan has been created, an ESN can then be implemented. Deploying new systems—including ESNs—can be problematic, so the organization’s strategic plan should be sure to address possible challenges, including the likelihood of employee resistance. Will employees adopt the new system? Not everyone uses every social media platform in their personal lives, so why should they use them at work? In order to ensure a successful implementation of an ESN, organizations can also follow industry best practices, or methods that have been shown to produce successful results in prior implementations. You’ll learn more about systems implementation in Lesson 12. When implementing an ESN, successful companies follow a process of four stages having the elements shown in Figure 9-13. Read through the items and reflect on what you went through when you first started using SM. Think about how important your friends were in your decision to start using SM. Having an internal champion or defender of your internal ESN is equally important.

Q9-7 How Can Organizations Address SMIS Security Concerns?

As you have seen, social media revolutionizes the ways that organizations communicate. Twenty years ago, most organizations managed all public and internal messaging with the highest degree of control. Every press conference, press release, public interview, presentation, and even academic paper needed to be preapproved by both the legal and marketing departments. Such approval could take weeks or months. Today, progressive organizations have turned that model on its head. Employees are encouraged to engage with communities and, in most organizations, to identify themselves with their employer while doing so. All of this participation, all of this engagement, however, comes with risks. In this question, we will discuss the need for a social media policy, consider risks from nonemployee user-generated content, and look at risks from employee use of social media.

Managing the Risk of Employee Communication

The first step that any organization should take is to develop and publicize a social media policy, which is a statement that delineates employees’ rights and responsibilities. You can find an index to 100 different policies at the Social Media Today website.42 In general, the more technical the organization, the more open and lenient the social policies. The U.S. military has, perhaps surprisingly, endorsed social media with enthusiasm, tempered by the need to protect classified data. Intel Corporation has pioneered open and employee-trusting SM policies, policies that continue to evolve as the company gains more experience with employee-written social media. The three key pillars of its 2020 policy, shown in Figure 9-14, are:

· Disclose

· Protect

· Use Common Sense43

 Figure 9-14: Intel’s Rules of Social Media Engagement

As with any relationship, comments can be inappropriate or excessively negative in tone or be otherwise problematic. Organizations need to determine how they will deal with such content before engaging in social media. This is done by designating a single individual to be responsible for official organizational SM interactions and by creating a process to monitor and manage SM interactions. This allows the organization to have a clear, coordinated, and consistent message. User-generated content (UGC), which simply means content on your SM site that is contributed by users, is the essence of SM relationships. Following are a few examples of inappropriate UGC that can negatively affect organizations. Problems from External Sources The major sources of UGC problems are:

· Junk and crackpot contributions

· Inappropriate content

· Unfavorable reviews

· Mutinous movements

When a business participates in a social network or opens its site to UGC, it opens itself to misguided people who post junk unrelated to the site’s purpose. Crackpots may also use the network or UGC site as a way of expressing passionately held views about unrelated topics, such as UFOs, government cover-ups, fantastic conspiracy theories, and so forth. Because of the possibility of such content, organizations should regularly monitor the site and remove objectionable material immediately. Monitoring can be done by employees or by companies such as Bazaarvoice, which offer services not only to collect and manage ratings and reviews but also to monitor sites for irrelevant content. Unfavorable reviews are another risk. Research indicates that customers are sophisticated enough to know that few, if any, products are perfect. Most customers want to know the disadvantages of a product before purchasing it so they can determine whether those disadvantages are important for their application. However, if every review is bad, if the product is rated 1 star out of 5, then the company is using social media to publish its problems. In this case, some action must be taken. Sometimes inappropriate social media content can come from unexpected places. In 2016 Microsoft released its artificial intelligence chatbot “Tay” on Twitter. Tay was supposed to increase user engagement by learning from them. Unfortunately, it learned to be extremely racist and sexist from its interactions. Microsoft disabled Tay after a series of horribly offensive tweets.45 Responding to Social Networking Problems Part of managing social networking risk is knowing the sources of potential problems and monitoring sites for problematic content. Once such content is found, however, organizations need to respond appropriately. Three possibilities in such situations are:

· Leave it

· Respond to it

· Delete it

If the problematic content represents reasonable criticism of the organization’s products or services, the best response may be to leave it where it is. Such criticism indicates that the site is not just a shill for the organization but contains legitimate user content. Such criticism also serves as a free source of product reviews, which can be useful for product development. For the criticism to be useful, the development team needs to know about it, so, as stated, processes to ensure the criticism is found and communicated to the team are necessary. A second alternative is to respond to the problematic content. However, this alternative is dangerous. If the response can be construed in any way as patronizing or insulting to the content contributor, it can enrage the community and generate a strong backlash. Also, if the response appears defensive, it can become a public relations negative. In most cases, responses are best reserved for when the problematic content has caused the organization to do something positive as a result. For example, suppose a user publishes that he or she was required to hold for customer support for 45 minutes. If the organization has done something to reduce wait times, then an effective response to the criticism is to recognize it as valid and to state, nondefensively, what has been done to reduce wait times. If a reasoned, nondefensive response generates continued and unreasonable UGC from that same source, it is best for the organization to do nothing. Never wrestle with a pig; you’ll get dirty, and the pig will enjoy it. Instead, allow the community to constrain the user. It will. Deleting content should be reserved for contributions that are inappropriate because they are contributed by crackpots, have nothing to do with the site, or contain obscene or otherwise inappropriate content. Deleting legitimate negative comments can result in a strong user backlash. In the early days of social media, Nestlé created a PR nightmare on its Facebook account with its response to criticism it received about its use of palm oil. Someone altered the Nestlé logo, and in response Nestlé decided to delete all Facebook contributions that used that altered logo and did so in an arrogant, heavy-handed way. The result was a negative firestorm on Twitter.46 A sound principle in business is to never ask a question to which you do not want the answer. We can extend that principle to social networking; never set up a site that will generate content for which you have no effective response! Internal Risks from Social Media The increased adoption of social media has created new risks within organizations as well. These risks include threats to information security, increased organizational liability, and decreased employee productivity. First, the use of social media can directly affect the ability of an organization to secure its information resources. For example, suppose a senior-level employee tweets, “Married 20 years ago today in Dallas” or “Class of 1984 reunion at Central High School was awesome” or “Remembering my honeymoon to Hawaii.” All of these tweets provide attackers with the answers to password reset questions. Once attackers reset the user’s passwords, they could have full access to internal systems. Thus, seemingly innocuous comments can inadvertently leak information used to secure access to organizational resources. Unfortunately, it turns out that it’s not a good idea to tell everyone it’s your birthday because your date of birth (DOB) can be used to steal your identity. Employees using social media also can unintentionally (or intentionally) leak information about intellectual property, new marketing campaigns, future products, potential layoffs, budget woes, product flaws, or upcoming mergers. It’s not just information leakage, either. Employees may install unauthorized apps that deliver content using SM that bypasses existing security measures. Or worse, they may use their corporate password at less secure SM sites. Second, employees may inadvertently increase corporate liability when they use social media. For example, suppose a coworker regularly looks at SM content with questionable sexual content on his or her own smartphone. The organization could be slapped with a sexual harassment lawsuit. Other organizations may face legal issues if employees leak information via social media. Schools, healthcare providers, and financial institutions must all follow specific guidelines to protect user data and avoid regulatory compliance violations. Thus, tweeting about students, patients, or customer accounts could have legal consequences. Finally, increased use of social media can be a threat to employee productivity. Posts, tweets, pins, likes, comments, and endorsements all take time. This is time employers are paying for but not benefiting from. Udemy notes that SM sites affecting employee productivity include Facebook (65 percent), Instagram (9 percent), Snapchat (7 percent), and Twitter (7 percent).47 From an employee’s point of view, you might think a little lost productivity is OK. But imagine you’re the employer or manager, which hopefully you’ll be at some point. Would you mind if your employees spend their days using SM to look for another job, chat with friends, or look at vacation pictures when your paycheck is tied to their productivity? What if SM is being used for interoffice gossip that creates HR problems, morale issues, and possible lawsuits? Smart managers will understand that, like any technology, SM comes with both benefits and costs.

Knowledge Check

Q9-8 2031?

Social media has hit a rough patch lately. Facebook founder Mark Zuckerberg was recently asked to testify before the U.S. Congress about a large data breach affecting more than 80 million people. It turns out a researcher had siphoned off Facebook user data and then sold it to an analytics firm before Facebook enacted tighter data restrictions. This violation of user privacy resulted in a swarm of news articles critical of all social media. Over the past decade, social media has been seen as something fun that could reach customers in new ways. And it was fun. It did reach customers in new ways that changed the marketing landscape. Social media companies received generous praise and huge market valuations. But there was always the question of how these companies were going to make money. Most of them relied on selling user data to increase ad revenue. The general belief was that users wouldn’t mind the “small” loss of privacy if they got the service for free. But what if they did mind? By 2031, the social media landscape will look much different than it does now. The honeymoon phase of social media is over. Privacy is becoming important again, teens are leaving Facebook,48 and quitting social media (or at least taking a break) is becoming a cool thing to do.49 In many ways, social media is like your driver’s license. It made driving fun and exciting when you first got it. But over time, driving became more of a utility. It became something you had to do to get from point A to point B. What happens when social media becomes a utility? What happens to social media companies when their products (you) decide to leave? Whose information will they sell? There are still tremendous opportunities for growth in the social media space. Enterprises are starting to use it internally (Enterprise 2.0). Is there an Enterprise 3.0 around the corner? New mobile devices with innovative mobile-device UX, coupled with dynamic and agile information systems based on cloud computing and dynamic virtualization, guarantee that monumental changes will continue to occur between now and 2031. (See Figure 9-15.)

Figure 9-15: Redesigning Enterprises for Social Media

Source: Rawpixel.com/Shutterstock

The explosive growth of IoT devices has opened up entirely new markets for social media. For example, a network-enabled fitness tracker can now send workout data to the cloud, where it can be used as part of a friendly competition with friends. Fitness trackers can now be part of a larger social interaction. Imagine the new types of social interactions that will come when mixed-reality devices become popular. You could sit in a virtual college class, play an online game with friends, and IM coworkers all while you’re physically sitting at your desk at work. Organizations like Harvard, Microsoft, and Starbucks are concerned enough with social media that they have hired chief digital officers (CDOs), a position responsible for developing and managing innovative social media programs.50 Think about your role as a manager in 2031. Say your team has 10 people, three of whom report to you; two report to other managers, and five work for different companies. Your company uses a variety of desktops, laptops, tablets, phones, and virtual devices. Some of these are issued by the company, but most are brought in by employees. All of these devices have features that enable employees and teams to instantly publish their ideas in blogs, wikis, videos, and whatever other means have become available. Of course, your employees have their own accounts on Facebook, Twitter, LinkedIn, Foursquare, and whatever other social networking sites have become popular, and they regularly contribute to them. How do you manage this team? If “management” means to plan, organize, and control, how can you accomplish any of these functions in this emergent network of employees? If you and your organization follow the lead of tech-savvy companies such as Intel, you’ll know you cannot close the door on your employees’ SM lives, nor will you want to. Instead, you’ll harness the power of the social behavior of your employees and partners to advance your strategy. So what, then? Maybe we can take a lesson from biology. Crabs have an external exoskeleton. Deer, much later in the evolutionary chain, have an internal endoskeleton. When crabs grow, they must endure the laborious and biologically expensive process of shedding a small shell and growing a larger one. They are also vulnerable during the transition. When deer grow, the skeleton inside grows with the deer. No need for vulnerable molting. And, considering agility, would you take a crab over a deer? In the 1960s, organizations were the exoskeleton around employees. By 2031, organizations will be the endoskeleton, supporting the work of people on the exterior. What all of this means for you is that social media + IoT + cloud will create fascinating opportunities for your nonroutine cognitive skills in the next 10 years!

So What? Evolving Social Media

Take a moment to think about your parents, grandparents, aunts, or uncles—how many of their childhood friends do they keep in contact with on a regular basis? Some of them likely have no contact whatsoever with friends and acquaintances that they met in early stages of life. On the other hand, some may be quite sociable and seek out these types of interactions, which begs the next question—how do your family members keep in contact with these acquaintances? It is highly likely that social media served as an important medium by which these long-lost connections were reestablished and by which they are now being maintained. Now consider your own social media use. When did you create your first social media account? Which platform did you choose, and why did you choose that one specifically? Was your choice motivated by the platform that your friends were using (you were pulled in), or were you motivated by finding a platform that certain people were not members of, like your parents (you were pushed out)? There are certainly many options to choose from today as more and more startup companies have joined the social media space. Interestingly, many of social media’s newfound success stories have been bought out by the existing juggernauts. For example, Facebook acquired Instagram for $1 billion in 2012 and WhatsApp for $19 billion in 2014.51 However, while the owners of these platforms may change, the form and function of social media platforms remain largely the same as they were many years ago. They are places for people to connect and share information, in the form of text, images, videos, and feedback (e.g., likes!). While there doesn’t seem to have been many innovations thus far in the world of social media, that doesn’t mean that things can’t change. Here are some predictions that have been made about what our social media experiences may be like in the future.

Source: NicoElNino/Shutterstock

Evolving Trends While the basic underpinnings of social media will likely persist, namely connecting people together and allowing them to share information with each other, the nature of that information exchange will probably be very different over time. For example, with the growing adoption of and potential for wearable technology, social media interactions will take place using these devices. Smartwatches are widely popular, and they can already be used to receive notifications and messages and even respond to social media posts. Companies are also developing and starting to market smart clothing, which is being designed to display notifications to wearers through buzzes and visual alert mechanisms. Another area beginning to experience growth is augmented and virtual reality. Augmented reality is rapidly gaining momentum in the workplace as the ability to superimpose imagery and information in the field of view of workers can add tremendous value. Consider the warehouse forklift driver who is given optimal routes driving through a warehouse, alerts of other forklifts operating in the area, and real-time order info on items and quantities that need to be picked up and delivered to the shipping portal. In a social media context, augmented reality can offer frictionless information sharing as users will no longer need to retrieve phones or tablets but can constantly have updates and posts delivered in their field of view in real time. Other evolving trends in social media right now52 include speculation that platforms may abandon the “like” functionality to remove power from influencers and restore power to the platforms, which are seeking to maximize ad revenues. Some experts believe that we may see a restoration of privacy in some platforms as users become more aware and savvier about the troves of data that are currently being collected about them. Additionally, the age ranges of users may shift as there are reports that younger generations (ages 12–34) are exhibiting stale numbers and that, in some cases, numbers are decreasing.53 Finally, expanding Internet connectivity around the world will continue to fuel the rise of smaller and more localized/specialized social media platforms that will compete with—but not replace—the existing powerhouses. The Fringe Experts have also made some intriguing predictions about the nature of social media a decade or two into the future.54 For example, one prediction is that wearables will someday evolve from something on our bodies to something inside our bodies that will be linked with our cognitions—instead of sharing just text or images, you may be able to share a taste or a smell. Another prediction is that standard input/output devices will be replaced by voice controls and holographic displays, rendering our handheld devices obsolete. One final prediction goes beyond how we share or receive information—rather, it focused on the nature of that information—one expert believes that all of the data we generate will be analyzed and used to create highly personalized experiences for each user, something that is still just in its infancy today. Questions

1. Consider and answer the questions introduced in the beginning of the article. When did you create your first social media account? Which platform did you choose, and why did you choose that one specifically? Was your choice motivated by the platform that your friends were using (you were pulled in), or were you motivated by finding a platform that certain people were not members of, like your parents (you were pushed out)?  Show Answer

2. How often do you engage with social media? As you read the article and learned about the emerging technologies that may be used as a part of future social media experiences, how did you feel? Are you excited at the prospect of augmented reality and wearables, or are you concerned that these technologies may be too intrusive or will be tiring as you can constantly be connected?  Show Answer

3. The article mentions that younger generations are starting to plateau in their use of social media—what do you think is driving this trend? Are you a part of this trend?  Show Answer

4. Do you think there are any adverse effects of social media use? Have you ever noticed any of these effects in your own life?  Show Answer

Security Guide

Digital Throne Of Lies When was the last time that you told a lie? Have you lied yet today? OK—enough about you. When was the last time that you caught someone else in a lie? Chances are high that you have engaged in multiple forms of deception in the past 24 hours, even if these were only benign lies with minimal consequences (e.g., texting a parent that you were working on schoolwork when you were really just hanging out with friends). However, the frequency with which you have caught others in deceptive acts is likely quite a bit less than your own deceitful behavior. Why would there be this discrepancy? Are you just a horrible person relative to those around you? Relax—that is not the case at all. It all comes down to your ability to identify deception. Scientists have exerted tremendous effort to understand deception as it is such a mainstay of human communication. A focal point of deception research is the investigation of detection accuracy rates—in short, scientists want to know how good we all are (on average) at identifying deception. A review of the countless studies that have looked at deception accuracy rates indicates that humans are only about 54 percent accurate—a hair better than chance. But why—why can’t we more accurately detect deception? As it turns out, there are many factors that have led to our inability to accurately identify deception. One theory is the evolutionary perspective, which assumes that only our distant ancestors who were able to successfully lie to conceal the location of food and other resources survived. In other words, all of us living today are the by-product of generations of expert liars who leveraged deception to ensure “survival of the fittest.” Another factor is the complexity of communication, which includes verbal, nonverbal, and linguist cues. Our brains have limited capacity, and we cannot possibly keep track or accurately interpret all of these many communication cues (e.g., the words we say, how we say them, tone of voice, body language, eye behaviors, etc.). A more optimistic but naive-sounding factor is that we simply have a propensity to assume good in others and to trust them. However, we now find ourselves immersed in a digital world in which many of our daily interactions take place online. What does this paradigm shift in communication, from the physical to the virtual world, mean for us and our communication? How does it change our perceptions and trust of others, whether it is following the feeds of social media influencers or our online interactions with businesses?

Source: Igorstevanovic/Shutterstock

There’s No Lie in Online. Wait, Yes, There Is ... Social media serves as one of the most common digital forums for us to regularly communicate with each other. Unfortunately, the limited face-to-face interactions in social media allow us to use it as a lens through which we can project a misrepresented view of ourselves to others. (This is often referred to as impression management.) One study investigating online daters found that over 60 percent of people lied about their weight, and another study based on over 2,000 responses reported that 40 percent of men misrepresented themselves online.55 Sadly, as we engage with social media more and more and use it as a baseline for gauging our own accomplishments, the warped projections of others can diminish how we feel about ourselves. An area of social media that is rife with people attempting to project positive impressions of themselves is the domain of social media influencers. These paid digital advertisers post images of themselves using products or having experiences in an attempt to persuade the rest of us to make similar purchasing decisions. Businesses have recognized the profitability of leveraging influencers as yet another stream of digital marketing to the point that estimates project $15 billion will be spent on influencer marketing in 2022.56 Furthermore, about half of participants in one study indicated that they had purchased a product based on the recommendation of an influencer. But have you ever stopped to think if you can actually trust the influencers and their product recommendations (i.e., have influencers even tried/used the product that they are trying to sell)? A majority of participants in one study indicated that they hope this is the case as they thought marketing a product without experience with that product would be deceptive and unethical.57 Distrust, however, is not just relevant to dating websites, social media sites, or social media influencers. Truth and deception are also pivotal factors for companies and how they engage with customers online. An analysis of consumer perceptions of businesses and social media revealed that two-thirds of participants held brands accountable for mitigating misinformation and, even further, that brands should proactively police and eliminate damaging information online.58 Additionally, distrust of businesses online has also led over half of one study’s respondents to proactively reduce the amount of information that they share with companies digitally. Overall, our online interactions provide countless avenues to communicate accurate and inaccurate information. Whether one person is misrepresenting themselves to another on a dating site or an influencer is selling a product to droves of followers without any basis for doing so, we are all being bombarded with questionable information every time we look at a digital screen. Just like our most distant ancestors, our digital survival may depend on our ability to accurately discern truth from deception. Discussion Questions

1. Take an inventory of your text messages or emails for the past day. How many times do you even marginally misrepresent the truth? Does your own track record support the observation that most people regularly engage in some form of deception?  Show Answer

2. The article reports that the human deception detection accuracy rate is right around 50 percent. The studies used to calculate this statistic are largely based on participants who were college students or average citizens. Do you think law enforcement or government/military personnel would have higher accuracy rates?  Show Answer

3. Are you guilty of using impression management tactics on your own social media accounts? Or have you found yourself impacted by the possible impression management tactics of others? Overall, does the information in this article make you want to reduce your use of social media or abandon it altogether?  Show Answer

4. You have probably heard of the polygraph machine—a device designed to measure physiological responses of interviewees while they are being asked a number of questions to identify if deception is present. While these devices cannot be employed in our online interactions, do you think new types of technologies to identify deception are being developed?  Show Answer

Career Guide

Source: Adam Young, RC Willey, Social Media/Online Reputation Manager

· Name: Adam Young

· Company: RC Willey

· Job Title: Social Media/Online Reputation Manager

· Education: University of Utah

1. How did you get this type of job? I started working at RC Willey in the marketing department while I was going to college at the University of Utah. I worked my way up and took on social media and other online marketing opportunities when RC Willey decided it needed to move in that direction. I worked hard, and I did what it took to make the people around me successful. Those people recognized my hard work and my dedication to making the company successful, and my dedication paid off.

2. What attracted you to this field? As I was going through my degree and working to pay my way through school, some of the skills that came naturally to me were in marketing and technology. While looking for career opportunities, I saw that technology was always going to be important and was constantly evolving. I’ve always liked new challenges, and this looked like a good starting point.

3. What does a typical workday look like for you (duties, decisions, problems)? Every day I go over the reports from the previous day. I also create new reports to help me understand the day-to-day trends. I use my marketing tools to respond to positive and negative feedback and reviews, and I work with management to address any concerns. I create content for the various social media platforms and schedule it to be posted. I also attend meetings with the marketing department.

4. What do you like most about your job? One of the best things about my job is that I get to help people. Not everyone is happy with their current situation, and I get to reach out and do what I can to take care of those individuals and see what the company can do to address their concerns. One of the best things is when someone completely changes their mindset about the company because I addressed their concerns when they thought nobody cared about or was willing to listen to them.

5. What skills would someone need to do well at your job? Being able to manage your time is an essential part of being a social media manager. You also need to be a good communicator. Being able to analyze data and create the story that the marketing team wants to tell is key.

6. Are education or certifications important in your field? Why? Most companies require an educational background in marketing and/or communications or something similar. Most are looking for experience as well. While other certifications in media buying, content creation, influencer marketing, and so on, are beneficial, they are not always necessary.

7. What advice would you give to someone who is considering working in your field? Do as much as you can to gain as much experience as you can before you graduate. A lot of companies are looking for experience. The more that you do now to develop those skill sets, the better you will stand out.

8. What do you think will be hot tech jobs in 10 years? I think that most everything is going toward the tech industry right now. I think some of the hot tech jobs in industry will be programming engineers and digital marketing in different capacities.

Ethics Guide

Life, Liberty, and the Pursuit of Not Being Rated The sound of the alarm on the nightstand had no effect on Patricia—she had been awake for at least 45 minutes and had just been staring at the ceiling deep in thought. This was not an uncommon morning for her as she would often run through everything she had to do that day in her mind before getting out of bed. She and her husband, Richard, had been living their dream for the past 10 years. They had finally found the courage to step away from the stability of the corporate world to start their very own bed-and-breakfast in Vermont. Having a business of their own had been something they had talked about since they’d started dating 35 years ago, and they had finally made it happen. Patricia walked downstairs and started setting up the breakfast area, keeping an eye out for the delivery trucks. She and Richard prided themselves on the quality and freshness of the food they served, so every morning they had dairy products delivered from a local dairy and pastries dropped off from a local bakery. Farm fresh eggs and bacon? They had those, too. They wanted to be known as the best spot in the country for fueling up before hitting a long day of skiing, hiking, antiquing, or maple sugaring. After almost a decade of top-tier service, they had created what seemed like an ironclad reputation—she could count on one hand the number of truly unsatisfied customers who had called and wanted their money back (and she was still pretty sure that the problem had been with these crazy people and not her service). But things had changed drastically over the past couple of years. With the relatively recent popularity of travel booking and reviewing websites, online reviews had become the lifeblood of new business.

Source: Ico Maker/Shutterstock

Make-or-Break Reviews Nowadays, it seemed like people couldn’t so much as buy a toothbrush without checking what other people had to say about it online. In the past, if someone complained about the bed-and-breakfast, she could give them a refund or offer them a discount on a future stay, and that always seemed to resolve the situation. Now, unhappy customers just posted a bad review online, and the people who took the time to post a negative review always seemed to be making a sport out of how cruel and damaging they could be. The hardest part was that she never had the chance to try to resolve the situation—she didn’t even know who they were! Initially, online reviews of the bed-and-breakfast had been positive. But the trouble had really started over the past few months. There had recently been several really harsh reviews posted about their business in a relatively short period of time. Patricia had done an analysis of her bookings for the past several years, and there had definitely been a drop in business this year since the critical reviews had been posted. What had been even more troubling was that she had carefully read the reviews and some of the details in each review just didn’t seem to add up. For example, one review talked about the “horrible time that their family had last weekend,” but Patricia checked the bookings, and there hadn’t even been a family staying with them during that time. Richard thought she was being paranoid, but Patricia’s theory was that a new, competing bed-and-breakfast in town was posting false reviews online about competitors in an effort to drum up more business for itself. The reality was that Patricia couldn’t take it anymore. She felt like she was fighting an invisible foe—no matter how positive her customers were about their experiences, negative reviews still seemed to crop up fairly regularly. She seemed to have no avenue for recourse, and she was fairly confident that the negative reviews were fake. She found herself constantly checking the online review sites, especially one in particular that seemed to be the most popular, and she had even gone to the trouble to email the site to see if she could just have their business completely removed from the listings. The good reviews were certainly helpful, but she just couldn’t take the bad ones anymore. She pulled her phone back out of her pocket and checked her email again to see if she had a response from the site, but nothing had come in yet. “How hard can it be to write an email?” she thought to herself. Just then, she was snapped into focus by the headlights of a delivery truck that had just swept across the driveway. “It looks like the eggs are here,” she muttered under her breath as she put on her jacket, opened the door, and started walking out to greet the driver. Reviews? You Mean Fake News Mark heard the familiar ping alerting him to the presence of a new email. He switched over to his email platform, looked at the message, and didn’t recognize the sender. Just before deleting the message, he quickly skimmed the title and realized that it may be a message he should check out. It was from some business owner complaining about too many negative reviews, and they even went as far as claiming that the negative reviews were fake—not an uncommon complaint these days. However, this person was wanting him to take their business off the travel review site that Mark helped run. He had to admit—these people had a point. In some backroom analyses the site managers were running on reviews, they found that it was common to see a lot of polarized reviews, either 1/5 stars or 5/5 stars, being submitted from the same IP address in a short period of time. The site managers knew that this was a sign of fake reviews, but hey, these posts kept the reviews fresh and increased traffic to the site, and no one could prove that they weren’t real. Furthermore, weeding out all of the fake reviews or removing businesses anytime business owners complained would cut out a large chunk of their Web traffic, which would be the kiss of death for any online company. Mark also thought to himself, “Hasn’t this lady ever heard of free speech? People can say whatever they want on the Internet—it is the Wild, Wild West!” He chuckled, deleted the message, and checked to see if he had any other email.  Discussion Questions

1. Think about Mark’s perspective regarding the fraudulent reviews that are posted on his company’s travel review site.

a. Is this behavior ethical according to the categorical imperative?

b. Is this behavior ethical according to the utilitarian perspective?

2. Do you agree with Mark’s opinion that freedom of speech means that people can post whatever they want on the Internet? If not, what can be done to govern the information that is shared on the Internet?

3. If you were in Mark’s position, would you remove Patricia’s business from the site? Why or why not? Explain.

4. Have you ever posted a fake review online, or has someone ever posted something inaccurate about you online? How did either of these behaviors make you feel?

Active Review

Use this Active Review to verify that you understand the ideas and concepts that answer this lesson’s study questions.

· Q9-1 What is a social media information system (SMIS)? Define social media, communities of practice, social media information systems, social media provider, and social networks. Name and describe three SMIS organizational roles. Explain the elements of Figure 9-3. In your own words, explain the nature of the five components of SMIS for each of the three SMIS organizational roles.

· Q9-2 How do SMIS advance organizational strategy? Summarize how social media contributes to sales and marketing, customer support, inbound logistics, outbound logistics, manufacturing and operations, and human resources. Name SM risks for each activity. Define social CRM and crowdsourcing.

· Q9-3 How do SMIS increase social capital? Define capital, human capital, and social capital. Explain four ways that social capital adds value. Name three factors that determine social capital and explain how “they are more multiplicative than additive.” Define influencer and describe how you could use social media to increase the number and strength of your social relationships.

· Q9-4 How do (some) companies earn revenue from social media? Define monetize and describe why it’s difficult for social media companies to generate revenue. Give examples of how social media companies generate revenue from advertising and charging for premium services. Define pay-per-click, conversion rate, and freemium. Define ad blocking and explain how it hurts online companies’ ability to generate revenue. Summarize how growth in mobile devices affects revenue streams. Explain why concerns about mobile devices limiting ad revenue are overreactions.

· Q9-5 How do organizations develop an effective SMIS? Discuss why aligning the development of SMIS with the organization’s strategy is important. Describe the process of developing an effective SMIS. List four common social media goals and describe why they are important. Define metrics, success metrics, and vanity metrics and give examples of metrics that could be measured for the four goals mentioned previously. Describe the importance of making personal connections with users.

· Q9-6 What is an enterprise social network (ESN)? Define enterprise social network (ESN) and describe the primary goal of an ESN. Define Web 2.0 and Enterprise 2.0. Explain each element of the model. Explain how changes in communication channels have changed the way organizations communicate with employees. Give an example of how an ESN could benefit an organization. Define best practices and explain how the ESN implementation best practices listed in Figure 9-13 could improve adoption of the ESN.

· Q9-7 How can organizations address SMIS security concerns? Name and describe two sources of SM risk. Describe the purpose of an SM policy and summarize Intel’s guiding principles. Describe an SM mistake, other than one in this text, and explain the wise response to it. Name four sources of problems of UGC; name three possible responses and give the advantages and disadvantages of each. Explain how internal use of social media can create risks to information security, organizational liability, and employee productivity.

· Q9-8 2031? Describe ways in which the use of social media is changing today. Summarize possible management challenges when controlling employees in 2031. Describe the text’s suggested response. Describe how a social media company might be able to benefit from an IoT device. Explain the relationship of the differences between crab and deer to this change.

Using Your Knowledge with iMed Analytics This lesson has given you several important models for assessing the iMed Analytics system’s social media program. You can apply the components of SMIS to understand the commitment that Dr. Greg Solomon and developers must make. You can use organizational strategy and social capital models to assess the desirability of social media to companies that participate in iMed. You can also consider whether iMed Analytics might want to generate revenue via a freemium model or by placing ads within the application. You can help craft an effective social media strategy for iMed Analytics and help Greg manage the risks of using social media.

Using Your Knowledge

· 9-1. Using the Facebook page of a company that you have “liked” (or would choose to), fill out the five components of an SMIS grid shown in Q9-1. Strive to replace the phrases in that grid with specific statements that pertain to Facebook, the company you like, and you and users whom you know. For example, if you and your friends access Facebook using an Android phone, enter that specific device.

· 9-2. Name a company for which you would like to work. Describe, as specifically as you can, how that company could use social media in each of the areas from Q9-2 listed in parts a through f of this question. Include community type, specific focus, processes involved, risks, and any other observations.

a. Sales and marketing

b. Customer service

c. Inbound logistics

d. Outbound logistics

e. Manufacturing and operations

f. Human resources

For a newly launched social networking company, having only 20 people sign up per day could indicate you are doomed to fail. This is especially true if the team of colleagues involved had worked for successful ventures like SocialNet and PayPal. However, in 2002, this was the case for LinkedIn, the business-oriented social networking site headquartered in Mountain View, California. Venture capitalist Reid Hoffman assembled the team of designers and engineers who created the social networking site.

Growing the Network

Growth for the company was very slow in the beginning. Some days, only 20 people per day would sign up to become members. Metcalfe’s Law states that the value of a social network is proportional to the square of the number of connected users of the system. By this logic, if LinkedIn couldn’t get users to sign up, the company wouldn’t be worth anything. However, the company finally became profitable in 2006 when it boasted more than 5 million members. In 2010, LinkedIn experienced hypergrowth and saw membership climb to 90 million, with more than 1,000 employees in 10 offices around the world.60 The company was adding nearly two new members every second! One year later, in 2011, LinkedIn celebrated its eighth anniversary and enjoyed a membership of more than 115 million members. That same year, LinkedIn became publicly traded on the New York Stock Exchange and was valued at more than $4.5 billion. At the time, LinkedIn was earning more than $150 million per year in advertising revenue, which was $15 million more than social networking giant Twitter was earning from advertising.

Unlike other popular social networks that focus on friendships and recreation, LinkedIn focuses on professional connections. People use the site to highlight career skills and promote professional résumés. LinkedIn allows people to be introduced via common connections and allows job seekers to become connected with people posting jobs. Employers can target the exact type of person they are looking for and post jobs for those qualified individuals to see. Connecting colleagues, professionals, recruiters, job seekers, and employers is at the core of what LinkedIn is all about. LinkedIn is for all people interested in taking their professional life more seriously.

In 2016, LinkedIn was acquired by Microsoft for $26 billion. At the time the company had more than 9,000 employees and claimed more than half a billion members in more than 200 countries worldwide. More than 20,000 companies buy LinkedIn Recruiter accounts to help find potential employees. Job seekers can comb through more than 11 million job openings posted by the largest companies in the world. The acquisition of LinkedIn raised a lot of questions. Why did Microsoft’s CEO Satya Nadella buy LinkedIn? Would the two companies be worth more together than apart? This was Microsoft’s largest acquisition, and it was coming on the heels of a disastrous acquisition of mobile phone maker Nokia for $7 billion.

Source: Jejim120/Alamy Stock Photo

In separate blog posts, Satya Nadella and LinkedIn CEO Jeff Weiner outlined the ways the two companies plan to integrate products and leverage Microsoft’s scale.61 Some of the ways they’re planning on achieving their shared vision include:

1. Incorporating LinkedIn’s identity and network in Microsoft Outlook and the Office suite

2. Supporting LinkedIn notifications within the Windows action center

3. Enabling members drafting résumés in Word to update their profiles and to discover and apply to jobs on LinkedIn

4. Extending the reach of Sponsored Content across Microsoft properties

5. Powering Enterprise LinkedIn Lookup by Active Directory and Microsoft 365

6. Making LinkedIn Learning available across the Microsoft 365 and Windows ecosystem

7. Developing a business news desk across the content ecosystem and MSN.com

8. Redefining social selling through the combination of Sales Navigator and Dynamics 365

Some of these ideas show real promise. They would allow the large core of Microsoft users much greater reach on LinkedIn. Nearly every corporation uses some form of Microsoft products. Its desktop operating system sits on more than 88 percent of all desktops. Within corporations, that percentage is even higher. If LinkedIn is successfully integrated across all Windows products, LinkedIn will have greater reach and more frequent contact with end users. Consequently, that could lead to rapid growth of its user base and its profitability. Questions

· 9-10. Why is growing the number of users such an important metric for social media companies? How does Metcalfe’s Law relate to the profitability of social media companies?  Show Answer

· 9-11. Most social media companies rely on ad revenue as their main source of income. What are other ways that LinkedIn generates income? Why is it important for a company to have multiple ways of generating income?  Show Answer

· 9-12. Why do recruiters and job seekers like LinkedIn? Explain why an employer may dislike LinkedIn. Is there a strategic disadvantage to having your employees list detailed profiles on LinkedIn?  Show Answer

· 9-13. LinkedIn targets a specific demographic: working professionals. They tend to be older and better educated. Why might advertisers be more interested in this group over others?  Show Answer

· 9-14. Microsoft creates software focused on supporting businesses. LinkedIn focuses on creating a platform for business professionals. Does the acquisition of LinkedIn make sense? What type of synergies could come from integrating products from these two companies?  Show Answer

· 9-15. How would the integration of LinkedIn into the Microsoft Office Suite be beneficial?  Show Answer

· 9-16. How would enabling draft résumés in Microsoft Word to update connected LinkedIn profiles be beneficial?  Show Answer

· 9-17. Suppose you are advising Microsoft about future acquisitions. Which company would you recommend as a good acquisition? Why?  Show Answer

Complete the following writing exercises