CIO Organization Memo Paper
John_matt
chapter_21.docxChapter Twenty-One: Service Management
Himanshu Shah
OVERVIEW
Information Technology Service Desk processes have matured and standardized over the years. Everything about service management can easily be found in the IT Infrastructure Library (ITIL) or in the documents of the International Organization for Standards (ISO). While these and others have become established as standards, the actual delivery of the service, the tools, knowledge management, centralization of service desk, and associated processes have evolved over time. They continue to evolve to ensure cost optimization, automation, and consistency in delivery of service.
In this chapter, we review the service management life cycle, understand the steps and tools required in delivering a service, and explore different models used for delivering the service.
SERVICE MANAGEMENT LIFE CYCLE
The service management life cycle is an iterative five-step process, as depicted in Figure 21.1.
Figure 21.1: Service Delivery Life Cycle
Service Modeling
The service management life cycle essentially begins by defining a service delivery model (Step 1) that can be a combination of following options, which we discuss in subsequent sections:
· Centralized or decentralized service desk
· Outsourced or in-house desk
· On-site, phone-based, or Web-based support model
· Multilingual options
· Generalist or specialized queue-based support
Service Delivery
Call (ticket) creation, categorization, assignment, and resolution are the activities of the transactional service delivery (Step 2) as managed by the support analysts. In subsequent sections we review the service delivery steps from call assignment until the resolution of the call.
Organizations striving to provide a service delivery differentiation over time need real-time engagement with customers and must consistently measure performance, identify opportunity areas, and enhance the service delivery mechanism through experiential learning.
Customer Engagement
Service delivery managers must have channels for real-time engagement, assessment, and escalation management. Engagement (Step 3) requires real-time understanding of the calls or requests in the queue and the ability to deal with peaks and troughs. Peaks in call volumes are dealt with by pulling support resources from other queues, requesting that resources that are not scheduled be made available on request, deprioritizing noncritical service support, or leveraging resources from other support areas. Doing this also requires an understanding of what drives the high volume. In some cases, interactive voice response (IVR), bulletin board, and plasma updates can be provided to effectively communicate about the issues at hand.
Engagement also requires periodically taking the pulse of service delivery by monitoring calls and interactions. Monitoring provides a greater understanding of the challenges in providing a great customer experience and insight into other opportunities for enhancing the service. Often customers request real-time escalations when their service delivery expectations are not met immediately. The ability to handle escalations, including the number of escalations and the associated resolutions, is a critical element of service delivery process.
Customer engagement through account managers, especially if the service delivery organization is off-site, helps customers understand the process, culture, geography, and knowledge barriers to providing effective service. In addition, receiving feedback based on resolved or unresolved calls using surveys provides a greater understanding of the delivery landscape. We discuss the feedback process in detail as part of the performance measurement process.
Larger service desk operations have specialist roles for managing customer communications. They provide updates on outages, performance measures, and larger process changes in the service delivery model.
Performance Measurement
Performance measurement (Step 4) of service delivery is closely connected with the culture of the organization. Process-driven organizations strive to pay more attention to hard measures of service desk performance. These include the number of tickets managed in a given period of time, ticket handling time, percentage of first-contact resolution (FCR), and the overall resolution percentage.
Higher handling time leads to increased FCR percentage as it gives the customer a greater opportunity to understand and resolve the problem at hand. However, it also increases the cost of service delivery and will increase the hold time of other customers in the queue who are waiting to be serviced. Increased FCR will certainly lead to a greater customer satisfaction.
Organizations driving purely for superior customer experience measure feedback on relatively softer measures. Softer measures include customer feedback surveys, internal assessments from real-time monitoring, and feedback from internal quality audits.
Customer experience feedback typically is gained through a formal survey process. In order to ensure that a fair survey is conducted, larger organizations outsource the feedback process to a third party. Others have an independent audit function that reaches out to customers independently. In either scenario, customers are asked to provide feedback on how their problem/query was resolved. They are also surveyed to understand how the call/service was handled behaviorally.
If the sample of customers surveyed is too large, the cost of feedback will increase and if the sample is too small, the analysis can lead to skewed interpretations. Selecting the right distribution of counts and variants of customers to be surveyed is a critical success factor of the measurement process.
Surveys can be conducted using form-based (electronic or paper) input. In some cases, interactive feedback can be provided at the end of the voice conversation. Answering questions with numerical scores provides only the statistical facts of the overall service provided. Personalized and commentary-based interactions for feedback can yield a greater understanding of the service delivery model and also increase customer confidence in the support delivery model. Interactive feedback sessions can be accomplished through on-site account managers or service delivery managers having a real-time interaction with customers. In addition, attention and response from executive management to critical feedback on service will further strengthen customer confidence in the service delivery process.
Most service delivery organizations have a quality and audit function that measures solutions and behavioral elements from a review of sample calls. Quality processes, on a real-time or off-line basis, can track process adherence as well as etiquette followed by the service delivery organization in its interactions with customers. With this approach, many service improvement opportunities can be identified internally, well before customers get to experience them. Quality checks and input also enable coaching of the IT analysts with the appropriate process knowledge and specific conversational improvements.
Service Enhancement
Service enhancement (Step 5) opportunities identified through soft and hard performance measurement processes provide the input into iteratively evolving the service delivery model, thus representing a complete and continuous cycle. Here are some examples of how the performance process can loop into identifying improvement areas:
· Recognition of those interactions where customers have had a better experience can be used to role model conversations.
· Resolutions on new call types should be fed back into the knowledge base. These will increase the internal knowledge and improve the FCR rate.
· Service delivery training programs built to provide initial and incremental training should directly correlate to the findings from the audit and quality processes.
· Automation and self-service opportunities should be pursued for those call types that are predictable and high in occurrence. Whether it is an IT system or access to Web systems, password resets should be driven through a self-service process rather than being manually reset by the service delivery organization.
Additional system-based enhancements can be provided, from tight integration of back-office systems leading to increased accuracy of customer profile information. The following examples illustrate integration possibilities in the back-office systems:
· Structured information management using a Lightweight Directory Access Protocol (LDAP)-like setup provides a strong integration with business systems and account management.
· Single sign-on access to multiple systems using a single password reduces end-user frustration from managing multiple accounts.
Standardization of software versions, hardware options, image management, and other controlling procedures also reduces the variants in technological environments, further leading to predictable service management and delivery.
Additional enhancements certainly can be driven following ITIL or a similar framework that includes:
· Configuration management to identify, control, and validate all physical assets, service-level agreements, and knowledge. This is also referred to as a configuration management database.
· Effective management of disruptions in the standard service offerings and restoration of the offerings to standard working expectations within the service-level agreements. This is referred to as incident management.
· Proactive management to prevent incidents from happening. This is referred to as problem management. It requires deeper understanding of known issues— expected failures, disruptions, and degradations—with the aim to have a plan to attack the root causes.
· Following a change management methodology to manage changes in the infrastructure using standardized methods and procedures. This provides an assurance that the changes will be made per previously identified schedule while reducing any reworking or duplication of efforts.
· Ensuring that the IT infrastructure is available based on service and cost expectations that organizations are willing to pay.
· Release of new changes to IT infrastructure software and hardware without disruptions.
· Qualification and quantification of IT service management as expected by the customers.
Continuous communication through reports on performance, trends, assets, and compliance.
WHAT IS SERVICE MANAGEMENT?
In an ideal world, the best-serviced customer is one who receives no service, assuming that the IT infrastructure and systems are seamless, fault free, and high performing. The next best option is where the faults are auto-corrected or a method exists where there is self-sufficiency and customers are IT aware and can resolve problems independent of the service desk. This is probably the reality that most CIOs dream of, but it is certainly a difficult goal to accomplish.
The process of effectively submitting, assigning, approving, solution provision (or resolving), and closure of a request coming into an IT organization can be called service management. How effective efforts are depends on the standards of a given organization and the criticality of the service delivery process in line with the organization's business needs. Most organizations measure service on the basis of critical parameters such as call volumes, handle time, resolution rate, and FCR. Other organizations use pure customer experience as the only performance measure. In reality, it is best to strike a balance between service desk measurement parameters and customer experiences.
The service desk serves as a single point of contact for users across all supported locations for all IT-related incidents, service requests, problem management, and change orders. IT customers typically use one of these channels to log their incidents or support requests:
· Self-service incident-logging e-mail tool
· Telephone call (typically a single internal number across geographies)
· Walk-in (in the case either of the other two options are not possible)
In the introductory paragraph, we mentioned that a lot of best practice scenario and details can be obtained from documented standards. As a result, here we focus on idiosyncrasies with service delivery, pragmatic understanding of real-life scenarios, and how organizations tackle the constant pressures from process management, cost optimization, and customer experience.
Receiving the Call
For the purpose of simplification, we use the terms service desk tickets and service desk calls interchangeably. It is of utmost importance to ensure that first-level ticket-handling staff members are IT generalists with operational awareness and good communication skills. We strongly emphasize communication skills in order to build a strong, centrally operating global desk. Later in the chapter, we touch on the challenges of dealing with an outsourced and/or remote service desk. With strong IT and communication skills, you will build a greater ability to resolve the call in the first attempt. If it is not resolved, you will ensure that the call is assigned to the right second-level (L2) support staff. Some organizations call the first-level staff L1 and others call them coordinators. Organizations striving for quick turnarounds and higher FCR rates tend to build technically competent L1 support models. The FCR rate of calls attended by L1 will be higher if those calls are relatively simple and routine, such as password resets. Chances of resolution during the first contact with IT have improved over time, based on some of these changes in support models:
· Evolution of technology has made it possible for the L1 staff to access end user workstations remotely and increase the FCR rate.
· Increased sophistication in knowledge management and self-service tools also increases the ability of end users to solve basic IT requests on their own. This does not necessarily increase the FCR. However, it does reduce the need to call for IT assistance in the first place.
· Categorization of tickets by the call coordinators along with an assurance that all pertinent request data are collected up front results in a faster turnaround time.
Figure 21.2 is a graphic representation of one support model that may be employed.
Figure 21.2: Service Management
Call Assignment
Once the call has been logged, if it is not resolved by the front-end L1 staff, it needs redirection. Redirection to specialist groups (L2) within the service desk or specialist groups (L3) outside in other departments, such as information security, server infrastructure, or network teams, can be challenging.
Some organizations prefer to have multiskilled L2 specialists within the service desk; others prefer to have L2 specialists purely for services like access provisioning or remote connectivity. In some cases, the L2 layer is relatively thin, and many of these requests are passed on to L3 resources in other departments.
Having a larger L2 population within the service desk can be cost prohibitive. Also, having all access rights for troubleshooting and resolution of tickets may not be possible. As a result, some service requests must be assigned to L3 groups. This certainly impacts on resolution time, as calls are now assigned to other departments.
L3 assignments and resolution depend on two critical factors:
· The assignment of calls to the correct resolving groups requires L1 and L2 staff with strong technology and business acumen.
· The operating-level agreements (OLAs) between the specialist delivery organizations and the service desk have to be very strong. A typical service-level agreement is set between the service desk and the customer groups to which the service is provided; OLAs exist between the service desk and the specialist departments that are internal to the IT organization.
Call Resolution
Service delivery organizations strive for a high FCR rate. FCR is measured as a percentage of calls resolved by L1 staff during the first contact compared to the total calls resolved by the service desk in a given period of time.
Measuring resolution time also requires effective tracking of call status. Resolution time is a typical measure of time difference between the time the service was requested and the resolution is provided, minus all the hold time required to obtain pertinent details from the requesting customers.
Modern-day service desk automation systems send out resolution messages (e-mail, Web, or text message-based update) to the serviced customers followed by an acknowledgment from the customers back into IT. The acknowledgment of resolution by end user customers moves a call from Resolved to Closure state.
We touched on what it takes to create a ticket, assign the ticket to various resolving groups, resolve the ticket, and close the ticket with customers. In the next sections, we review other elements of service delivery like call management, service catalog, knowledge management, and asset management.
Call Management
Service requests for password or user management can be managed either using a self-service mechanism or some degree of automation based on employee events. For example, creation of accounts for new hires, closing accounts for terminated employees, and changing access privileges based on job changes are closely tied with the human resources (HR) system. The changes in the HR system should be recognized automatically to ensure quick changes in account privileges. Other service requests that do not require special handling by the L3 organization can be managed within the service desk, and ones that require changes to the infrastructure, such as firewall configurations, network changes, and application changes, are redirected to other support groups.
Categorization of a ticket as a service request or an incident should be managed by the call-handing staff. IT incidents impacting wider customer population often result in a lot of tickets (calls) created for the same problem. Management and closure of these issues can be expensive if not handled appropriately. In such cases, all the individual tickets need to be treated as child tickets and should be closed with a cross-reference to a single parent ticket. The parent ticket typically must be addressed by a specialty L3 organization followed by a root cause analysis of the issue and an update to the knowledge base with relevant resolution details. In addition, in order to reduce the calls coming into the service desk, these situations require a broadcast knowledge of the current situation. Alerts can be provided on plasma screens, dashboards, intranets (web sites), e-mails, and text messages based on the nature of the problem and the organization's approach in communicating such alerts. Also, if these issues are related to outages in the systems or the network, they require close looping with the support organization. There must be a cumulative accounting of outages and assurance that operational-level agreements are managed with supporting functions using these details. In many cases, the problems might not be outages. They can be degradations in system performance from increased usage, increased transactions, or other limitations with infrastructure horsepower or network bandwidth. Issues like these also require close looping with the supporting organizations to track root causes and assessment for ramping up of network bandwidth or infrastructure capacity.
Service Catalog Management
In simple terms, service catalog management is a menu or catalog of services provided by the IT department. It essentially captures different services, categories, approval details, cost information (if applicable), and expected turnaround time for resolution or service. Turnaround time is one of the critical performance parameters for the service delivery organization. It is published to customers being serviced and tightly coupled with the service desk processes and OLAs with other delivery organizations. A typical list of services is illustrated in Table 21.1. A sample entry for e-mail service in the service catalog is illustrated in Table 21.2.
Table 21.1: Typical List of Services
|
Service Type |
Provisioning Medium and Processes |
Measurement |
|
Communication |
Mail, messenger, and broadcast services |
Uptime, capacity, and performance |
|
Desktop management |
Configuration and upkeep of standard workstation image |
Patching, compliance and performance |
|
Productivity tools |
Document, spreadsheet, presentation, and process management tools |
Patching, integration, and performance |
|
Network access |
Access to organization's systems over local, metropolitan, and wide area networks |
Uptime and performance |
|
Remote access |
Remote access mechanism for troubleshooting, knowledge sharing, and presentation |
Authorization, authentication, and availability |
|
Internet/intranet |
Restricted and unrestricted access to authenticated and white-listed domains inside and outside of the organization |
Uptime and performance |
|
Identity and access management |
Managing users, their profiles, personalization, and ease of access to business systems through single-sign-on configuration |
Authorization, authentication, availability, and self-service |
|
Threat management |
Protecting business systems from vulnerabilities and attacks |
Infections, threats, vulnerabilities, patching, and compliance |
|
Backups and archiving |
Ensuring high availability of information from disasters and data losses |
Backup success, recoverability, and restoration |
Table 21.2: Sample Entry for E-mail Service in the Service Catalog
|
Service Name |
|
|
Service Availability |
24 hours/7 days a week |
|
Eligibility |
Full-time employees and authorized contractors, standard desktop access, andactive employment status on corporate HR management system |
|
Service Creation |
Automatic creation on the basis of employment entry and inactivation upon termination of employment services as recorded in the HR management system |
|
Service Support |
Service desk through published contact mechanisms |
|
Cost |
Charge-back component of IT services |
|
Service Availability |
99.9% excluding scheduled outages every other Saturday from 12 midnight until 3 a.m. |
|
|
Account creation and updates within 48 hours of employment status changes |
|
|
Quota/capacity based on designation and specialized needs |
|
|
Restoration based on severity of outages |
Other services provided in the catalog can include extension to the basic IT services, such as application services (support processes for enterprise resource planning and customer relationship management systems) and specialized services for projects and technology architecture management.
Knowledge Management
Service desk knowledge is not the same as process expertise within L3 groups. Service desk knowledge is built, managed, and accessed using a knowledge database. This database can be a simple collection of files, a repository on the intranet, or a systematic solution that provides a database of known issues/problems along with the resolution steps describing underlying actions. These knowledge tools provide details on IT issues along with resolutions. Organizations seeking customer excellence and process optimization tend to implement high-end knowledge management solutions. Organizations that have cost as the primary driver build rudimentary knowledge bases that are hosted on their intranet.
Sophisticated knowledge management solutions allow creation of structured knowledge with an easy-to-use unstructured search mechanism for service desk staff or end users. In most cases, resolution process/documentation is a narrative presentation that allows easy-to-follow step-by-step instructions for resolution.
Whether a knowledge management solution is bought or a rudimentary knowledge base is built in-house, service desk performance will depend on how the knowledge is maintained in the system. Every time a new call type is logged and resolved, it is imperative that the new knowledge (resolution) is incrementally added to this repository. This discipline will positively impact long-term resolution rates (FCR) and will ease the analysis of root causes.
Asset Management
Many service requests result in allocation of assets to end users in the form of software or hardware. It is essential that when IT analysts serve these requests there is a close tie-in with the software and hardware inventory management. The inventory will certainly tie into the organization's security, data protection, software compliance, and audit policies. Also, it is relatively easy to track and manage inventory of hardware resources as they are physically allocated. If these are devices connected on the network, inventory and desktop management software solutions can easily account for these resources and help you update the inventory automatically. In the case of software installation on desktops, installed units can be maintained as part of the inventory, and new installations can be made using centralized desktop management software with a push mechanism. Desktop management solutions can actually record the number of available licensed software units, track the number of installed software units, and appropriately lock from deployment unauthorized or unlicensed installations. Automating these inventory details improves accuracy of information, reduces the effort required from HR, and protects the organization from licensing violations.
Asset inventory can be managed through a sophisticated asset management tool or a simplified list of assets in the inventory. A typical asset inventory would capture these details:
· Asset identification number (internal to organization)
· Asset name and description
· Asset type (software, desktop, server, etc.)
· Asset location and assigned to user
· Manufacturer's serial number
· Asset make
· Asset model (if applicable)
· License key required
· Asset purchase date
· Asset install date
· Warranty/maintenance period
SERVICE DELIVERY MODELS
Economies of scale, particularly as cost optimization becomes an important driver for many organizations, can be generated by centralizing service delivery organizations in one or two locations. Language, culture, and time-zone differences complicate the centralized service delivery model for those organizations that operate in multiple continents or countries.
Centralized or Decentralized
Centralization gives companies the option to operate the service delivery organization from locations that are business critical, talent rich, and cost effective. It also allows for planning of shadow or backup HR, easy sharing of knowledge, service desk specialists (L2) sitting right next to the L1 service delivery center, and the ability to support multiple time zones from a single location. In addition to offering cost advantages, it also results in tight coupling of IT's people, processes, and technologies.
The next critical success factors should be a focus of service delivery in order to achieve excellence from centralization:
· Greater understanding of the customer organizations' culture by the servicing geography
· Acknowledgment of process, technical, and cultural gaps in the servicing geography
· Continuous improvement road map to bridge gaps periodically
· Strong communication channel and a continuous exchange of knowledge and information
The other opportunity lies in ensuring that there are designated account managers in the major geographies that are being serviced. They provide a continuous communication channel to the service delivery organization with knowledge regarding user expectations, cultural awareness, changes in technology, and other challenges in service delivery.
It is practically impossible for multilocation organizations to have a local service desk at each geographic location. In some cases there may not even be an IT presence at remote or smaller sites, which are typically satellite offices connected to the corporate network but with a small number of employees—sales offices, for example.
Decentralized service desks are built to provide in-person assistance. They are also implemented when support is required in local languages.
Outsourcing
Another popular mechanism of service delivery is a combination of centralized, global, and outsourced service delivery. Outsourcing enables organizations to leverage third parties that have built delivery models and other strengths from years of experience. It also allows scale-up and scale-down of resources and delivery models based on seasonal and other time-sensitive requirements.
Outsourcers bring models, tools, workforce, and processes with a variety of options. Options available with outsourcers may not readily fit into the needs of the organization being serviced. Critical decision points to be considered include:
· Identification of performance measurement metrics. These require fine-tuning over time, as the previously identified measurement parameters based on contractual agreements may not be strongly connected with the real-time customer experience. In an effort to drive process and performance, third parties may dilute the customer experience.
· Tools and processes for ticket management, change management, asset management, and knowledge also require fine-tuning.
· Communication and interaction processes with customers need to be built, validated, tested, and refined over time. Script-based interactions offer a seamless structure but also create a box that might impact customer experience and limit creativity of the service desk staff.
· Outsourcing the service to a third party will require a definite investment to bring the third party closer to the culture of customer organization. If the service is outsourced to a third party in another geography, additional investment and training will be required in building soft skills, customer service etiquettes, and understanding of cultural differences.
Building rapport with customers is yet another piece of the puzzle in real-time interactions. Rapport building can increase the call-handling time and will impact the overall service desk performance. In the case of real-time interactions, to simplify rapport-building conversations with customers, organizations may broadcast updates on events in the customer's location. These broadcasts educate service desk staff about major news, events, weather details, celebrations, and sporting events at the customers' sites.
Service Delivery Medium
Real-time interactions provide a greater customer experience, but they come with their own price tag. Evolution of technology has provided options to increase self-service usage through knowledge bases, IVR, and other tools for account password resets. These are the alternatives, in order of decreasing customer interaction and increasing self-service capabilities:
· Regarding service delivery, the first choice is to have an IT analyst walk to your desk. This method works very well. It allows easy rapport building as well as addressing issues other than your primary request. This is a luxury and not a very pragmatic solution as the cost of doing so is very high.
· The next best option is to make available phone-based support where you still have a remote but real-time conversation. Customers call into a service desk phone queue. For seamless access from any of the offices, it is desirable to have a single extension to call across geographies. Although this is a personalized delivery of service, it does not allow the IT analyst to attend to multiple customers at the same time. In cases when there are multiple queues of servicing organizations, the phone systems are coupled with IVR and computer telephony integration (CTI). IVR allows customers to provide basic information about their query so that the call can be forwarded to an analyst skilled with the specific query type. This increases the probability of FCR. CTI is a simple integration of customer profile information on the service delivery systems with telephony. Customers input a recognizable code during the dial time so that all relevant information about their profile and history of existing reported but unresolved problems is automatically populated. This results in reduced handling time and enhanced customer experience.
· Web-based interaction is the next available option. This is in real time and interactive. It also allows some degree of multitasking for the service delivery organization and makes exchange of information easy. Also, some of the challenges with conversational skills, accent, and language barriers are diluted with this approach.
· With the evolution of the Internet and intranets, knowledge is easily accessible from the workstations. There is no service desk personnel engagement, but this method requires end users to dedicate time. Even though staff engagement does not exist, the content presented to customers can be personalized based on nature of their jobs and queries that they are trying to resolve. When a particular query is searched, the resolution can be in the form of sequential steps to walk end users through the solution. In some cases, solutions are built to detect the problems automatically and fix them based on known solutions.
In an absolute cost-conscious setup where the volume of requests is very high with a high degree of variation, organizations use e-mail as a medium. Turnaround is not as effective and not as personalized. In these situations, the service is typically outsourced, and it is probably the most inexpensive service delivery option. Customer experience can get compromised with this approach.
CONCLUSION
Organizations can choose from several models and tools for service delivery. To pick the right option, the criticality of providing a certain set of services, nature of the business supported, cost of providing the actual service, and impacts from centralized, offshore, and outsourced approaches must all be considered.
