Individual Experiment Topics

XYZ Fast Prototyping MGMT 3405

1

Definition – Fast Prototyping Fast or rapid prototyping is a methodical exploration of innovative concept(s) by quick assembly of pieces either tangible or intangible to validate assumptions which are important to implement the concept. The outline of this concept is described in article “Intuit Inc. Project AgriNova” published in HBR by Thomas Eisenmann and Tanya Bijlani. Quickly identifying & rapidly developing solutions for part of the system which could be potential road blocks is key to ensure success of the product. This does not need complete development of all (or some) parts

Problem Statement Our organization is specialized Business Analytics and Data Management expertise. Among other things, one of the requests often made by our customer is to give guidance on suitability of tool (or set of tools) for a particular task. Even though this knowledge is available within our organization, it is dispersed as the consultants are working with different customer.

We set out to address the issue of - timely availability of comparison metrics across tools - continuous update to the metrics being used to compared

After discussing with our executives we decided to build a web based application internally so that we can feed in the comparison data on continuous basis without spending too much time on reconciliation efforts.

There were few challenges to be resolved while addressing the issues given in problem statement. We conducted a brainstorming session within our organization. The outcome of this session was a list of important components outline of which is as follows:

- User Interface: The UI should be easy to use and intuitive enough to hide the complexity underneath. Unless the tool is easy to use people will be reluctant to use it.

- Data Update: The data should be fed in on continuous basis to ensure updates for the tools to be compared are captured on regular basis. If the data is stale it will raise the credibility issue of the presented comparison. We cannot compare data of outdated version of the tools.

- Contextual Text Mapping: The biggest issue is contextual mapping of text which describes a particular feature of tool, product or application.

Of course this list is not comprehensive, but we need to address these points to ensure viability of the entire efforts.

I think using “Fast Prototyping” to validate the feasibility of the components is best course of action before attempting to build this product.

Leap of Faith Can we build & expand? As the data volume increase the methods employed, especially the algorithm employed will perform satisfactorily? We decided to find this out. Can we win? We did not spend great amount of time with user experience. We took a leap of

faith by assuming that the team who participated in building UX is representative of future

users. I think we should be able to tweak UI based on usage analytics and statistic.

XYZ Fast Prototyping MGMT 3405

2

Experiment Outline

We addressed each challenging component of the system to ensure we are certain that it is worthwhile to spend time and resources on this application. The three main challenges listed above are entirely different from technology perspective. The method employed for each of the component of the system is also very different and is outlined below:

User Interface: Nothing beats rapid prototype building

while getting buy in from end users. First we designed the

user experience by giving a name for each of the action as

shown in the diagram. We outlined the tentative look and

feel of each screen, where it was important for us to point

out key element of this prototype. Next we built couple of

screens which tested actual user interaction using the

rapid prototyping tool such as Balasamiq. This step helped us to understand how end user

would use our tool. We gather typical sequence followed by our users. One of these captured

sequences is given on next page.

Data Update: Prototyping this was bit easier and quicker. There are several tools available for web crawling. We implemented Apache Nutch in open source community.

Contextual Text Mapping: The hardest parts of this prototyping effort. We set out to research on the latest in “Pattern Matching” on Google Scholar. We realized that there are three distinct type: Simple Character Mapping, Word or Token Mapping and Context based Token Mapping. We agreed that type c would be appropriate for our method. More research revealed research done by Yoav Artzi andLuke Zettlemoyer documented in paper titled “The University of Washington: Semantic Parsing”. (Link)

We quickly prototyped the algorithms mentioned in the document “Semantic Parsing”. We are evaluating the results of this prototype.

Conclusion

Key take-away from the learning on topic of “Fast Prototyping” for me:

- Figure out how to break down the topic into manageable areas for experimentation.

- Search for previous referenced and reviewed work using Google Scholar for each of the

areas.

- Most important: Win executives confidence early, that we can deliver, by demonstrating

key concepts with fast prototyping

XYZ Platforms as innovation outcomes MGMT 3405

3

Definition – Platforms as Innovation Outcomes

The platform is extremely powerful ecosystem that quickly and easily scales, morphs and

incorporates new features (planks). It can also quickly add users, customers, vendors &

partners. The platform is broadly divided into three broad categories: business model platform,

industry platform, technology/product platform. This concept is outlines in “The Evolution of

Platform Thinking” by Michael Cusumano.

Problem Statement

Comparing two tools or applications is challenging especially if you have difficulty gathering

comparable features. This same concept is true for services or products

as well. Take example of ATT U-Verse GigaPower VS Comcast xfinity.

Both of them claim to have fastest networks. It doesn’t call out cost of

ownership over the time, drop rates, Quality of Service, Experience or

Engagement. ATT as well as Comcast only publish metrics on their site

which is favorable. Ford and GM trucks are superior to each other with

outstanding post sales service. Samsung SMART TV VS Sony Bravia is

another example which “Outstanding” picture quality, with “Intelligent”

processor having “Higher” Hz confuses people. Daily brands like Coke

and Pepsi also claim to have superior test to each other.

I think there is scope to develop a “Platform” which will serve as a mean to compare tools,

application, services or products in unbiased way.

Leap of Faith

This concept is very expansive and it is very difficult for me to wrap my head around all the

leaps of faith that I have to take. Here is list of few that I could think of:

- People will really need platform to “Compare” before taking decision. Survey or small

application can answer this leap of faith

- The technology to bundle this entire thing together can be built in orderly manner so as to

meet expectations of the user base. This will be very complex setup since there are several

complexities involved in this platform this will be very complex setup.

- The financial model will be viable and attractive to investors and end users. I will have to

build financial model and spend some time tweaking it.

Developing Platform Concept

I think problem stated above can be best addressed by

building an extensible platform. This is very good case

for business model platform + technology platform.

Please note that even though this platform serves

industries, it is not a “industry” platform as defined

paper sited previously To draw similarity between the

concepts taught in the class and in book “The age of

Platform” by Phil Simon, we have platform called as

“Comparison” platform (in blue) which consists of various module such as Technology,

Comparison Platform

Technology Adoption Verification

Autos Commodities IT Services Telecom Services

Add new feature to existing tool/service/product

Add new industry

XYZ Platforms as innovation outcomes MGMT 3405

4

Adoption initiative such as Marketing & Verification mechanism to ensure the validity of the

feed and output. Various businesses (industries) can be aligned for common feature

comparison.

Please note that there is subtle difference here as why they are called business model rather

than industry. The comparison business model for each industry will be different and hence it

will fit in as “Business Model” for that industry in the platform that I am proposing to develop.

To give you example, the commodities (e.g. cloth, office supplies etc) are best served with “Ad-

supported” business model. While the “Business Model” for auto industry could be

“Commission” based and for IT services could be “Pay as you Go”. Of course I agree that these

terminology need to evolve as we think more about what can be accommodated in this

platform, but this could be draft 1.0 terminology, if I can call it so.

Going forward the features could be added to continuously engage end users and add to the

existing user base. We can roll out platform with possibility to add new features for predefined

industries, but the platform con be expanded with possibility to add entire industry to it. Of

course making it available on multiple devices such as laptop, tablet, phone, various wearable

devices could be akin to adding “different” types of planks to expand the user base.

Experiments

To answer our favorite questions: Is it real? Can we win? Is it worth doing? I conducted couple

of experiments as follows:

- Is it real?1.0 Information overload: Tried google/bing/yahoo/wolfram alpha/duckduckgo

search for “Google Nexus 5 VS Samsung S5 VS iPhone 5S”. The search engines confuse

themselves at best and at worst present information which is way too much for a person to

quickly decipher.

- Is it real?2.0 Anybody out there: People are aware of this problem and trying to solve it.

Please take a look at www.versus.com. They are building exact same thing that I am

proposing. Other examples are Burstorm and Passioned Group.

- Can we win? Then why I think we will win: What people have built so far, included versus is

not comprehensive platform where we can add planks to extend it indefinitely. This is a web

application, the best one to be found though. Yelp and other user review sites are “Review”

site, they tell us how the product experience (perception sometimes) is, but do not

compare the features, which is critical for informed decision. This is definitely a green field

for our platform. Similar by limited application in medical field can be seen in this paper

- Is it worth doing? Service is where the returns are: I think it is worth doing it, but need to

conduct additional experiments to study financial feasibility.

Conclusion

I feel the concept presented here is a classic case for building a platform. It serves well to the

various dimensional attributes such as working in ecosystem, collaborating for usage and

collaborating for marketing along with technology sharing. The “Business” planks to this

platform can be added based on user’s demand, while “Technical” Planks can be added as

usage grows.

XYZ Open Innovation MGMT 3405

5

Definition – Open Innovation

The outline of open innovation can be best described by the concept of funnel shown in the

picture picked up from paper “Renewing growth from Industrial R&D” published by Prof Henry

Chesbrough. This image has been edited

to support the concept which we

implemented within our organization. The

external experts (red bean) contribute to

concept that is supported by the team of

internal experts (blue bean) who are

funded by organization to address the

value proposition for current customers.

The exchange of ideas and concepts along

with developed artifacts benefit all the

players. Of course sponsoring

organization has the “early mover” advantage. It also gains enormous clout over the direction

of research and development. Additionally it builds loosely integrated network of professionals

in the same domain which helps grow each other’s business.

Problem Statement

Technology especially the one developed in Silicon Valley has undoubtedly changed the face of

many industry. Shipping industry for example uses technology such as advanced optimization

algorithms to utilize every inch of space on ship and in container. Ships efficiently plan routes,

ports and take advantage of sea currents and weather conditions.

Medical industry hasn’t seen such advancement towards optimization yet, despite itself being

founded on one of the most talented community members. With explosion of Social, Local,

Collaborative platforms I think time is ripe for medical industry to take advantage of “Open

Innovation”

Some of the implications when you put these two things together:

- Medical professionals do not adapt information technology with same zeal as other

labor intensive bodies of our society such as shipping industry. Main problems for this

behavior being “Privacy Issues” associated with medical care and extremely

personalized nature of diagnostics. Secondary problems are lack of interconnected

backbone due to rules and regulation as laid out by HIPAA along with ambiguity of

description of cause and real reason behind the cause

- This makes it difficult for patient to discover alternatives for their problems

- It is also difficult for medical facilities such as hospital, doctors and nurses to discover

what their patients (a.k.a. customers) physical and emotional reactions are to the

provided care.

“Open Innovation” is the best means to bridge the gap between technology companies like ours and

medical professionals across the globe who are willing to participate in these efforts.

Ref: Renewing growth from Industrial R&D Prof Henry Chesbrough, UC Berkeley

XYZ Open Innovation MGMT 3405

6

Experiments

To solve this problem our team executed series of experiments as below:

1. Formed internal team (blue beans in the picture above) of “so-called” domain experts.

The reason I call them “so-called” is: they are all engineers and none of them has any

medical experience. They have little exposure to medical terminology since they are

exposed to IT work in medical environment. They are extremely passionate to make the

difference in the world.

2. Armed this team with tool to put the thoughts across to the world. This team rapidly

built an application to collaborate with willing experts.

(http://healthcaredataanalysis.org/)

3. We reached out expert within the industry to discover real “experts” who are respected

in this field (red beans in the picture above) to supplement our internal experts.

4. We produced the results & appreciated this open community members to contribute

Outcome & Result

Even though this effort started and sustained purely by passion to “Do good in return”

philosophy, to support it on an ongoing business we needed to show outcome and result to our

management. Here is list of few outcomes:

- Real help to people in need to understand medical care and optimize their spend by

relying on right resources (doctors, hospitals and medicines)

- Possibility to expand to other social media (currently this

supports only twitter)

- Collaborative network of over 100+ real industry experts

who can be contacted for additional business

opportunities

Here is list of few results:

- Actually ended up winning us business for one of the

largest chain of hospital in Arizona.

- Was sold as value added service to medical facility in bay area.

- Repository of analytics tool (not the diagnostics by contributors) which can be sold as

value added service.

Conclusion

I learned a great lesson from observing these passionate “Open-Source Nerds” learn and

experiment with “Open Innovation” to bring value to community. I also noticed the acumen of

managers to challenge and channel the spare energy available within system for future benefit

of organization. Couple of unaddressed questions which sprung to my mind: Whose intellectual

property it could be? How people contributing to it should be compensated? Are they

acknowledged appropriately? I guess the “Open Innovation” is driven by passion and tendency

of humans to “Do Good”, hence some of the questions could be irrelevant. None the less, this is

powerful tool in the hands of managers for innovation.