building a data & analytics product business

13
Building a Data & Analytics Product Business The Role of a Data Product Leader Sumit Dhar September 18, 2015

Upload: sumit-dhar

Post on 16-Apr-2017

487 views

Category:

Data & Analytics


0 download

TRANSCRIPT

Building a Data & Analytics Product BusinessThe Role of a Data Product Leader

Sumit Dhar

September 18, 2015

Table of ContentsThe motivation is to elicit interest in the emerging area of Data Products and explain how this may be of interest in the Analytics Services industry

1. The Big Data & Advanced Analytics Market

2. The Data Product Paradigm

3. Opportunity for IT/ITeS Service Companies

4. Considerations for building Data Product

5. Graduating B2B Data Product Ideas

6. Data Product Enabled Analytics Delivery Model

7. The Role of a Big Data Products Leader

8. Postscript…

9/18/2015 © Sumit Dhar 2

The Big Data & Advanced Analytics Market

o More than 2.5 quintillion bytes of data are created every day through a range of activities including social media, purchase, cell phone GPS signals, SCM/ logistics, videos, pictures, and audio

o Global spending on big data (software, infra, services etc.) is growing at an average annual rate of nearly 30 percent and is expected to reach $114 billion in 2018

o Successful leaders are springing forward with analytics through technology and organization framework that support rapid experimentation and innovation to evaluate risks and tradeoffs, understand cost and revenue drivers, and predict trends to help drive business performance and innovation

o However there are considerable barriers:o Talent gap; 4.4 MM trained man power need in Year 2015 alone

o Investment in data ingestion outstrips that required for analysis and harnessing in business outcomes

o Many organizations are struggling with insufficient systems

o Targeted Data Products developed to solve specific business problems can be used to embed high quality prediction and advanced analytics in business processes in a cheap, manageable and repeatable manner

Big Data and Advanced Analytics hold great promise for the global economy; Touted as “New Oil”; Barriers create new opportunities for Productized delivery

9/18/2015 © Sumit Dhar 3

The Data Product Paradigm

oAny App, solution, software, online component- that consistently generates or consumes DATA in a manageable and repeatable manner is a Data Product

- Term originally coined by DJ Patel erstwhile of Google

oData Product is a productized and engineered framework to ingest data and provide predictions, recommendations and insights for specific business use cases

- Not any single software

o The basic premise is to build scalable Big Data solution that can be used to federate best in class Data Science work product for specific Business Use Case across Businesses

- Reduce variability through design- Single solution with strong service layering through a SaaS based Analytics service- Manage need for differentiated competency for Advanced Analytics through use of SaaS

Analytical Services companies can conceptualize and leverage Data Product in helping large corporation to benefit from Big Data and Data Science in a scalable manner

9/18/2015 © Sumit Dhar 4

Opportunity for Analytics Service Companies

• Augment client abilities to quickly monetize emerging business areas utilizing Big Data & SMACMarket Penetration

• Engineering and Science driven “productized” analytics delivery engine

• Scale to solve specific business problemsScale

• High Quality Advanced Analytics work product made available at scale without the associated build cost for each initiative

• Optimization and Business Integration to be done as a front end work client management task

Non-Linear Growth

• Greater cloud based technical integration with client system to be able to drive positive changes through quick decision and action

• Payout based on incrementality not on FTE costOutcomes

• Potential opportunity to create IP based on Discovery ActivitiesCreating IP

“Productized” Delivery helps in improving scale, targeting nascent area of capability and build scale

9/18/2015 © Sumit Dhar 5

Considerations for building a B2B Data Product

Data Ownership

• In B2B space customer, or ownership and usage rights are sometimes uncertain

•Solution with anonymized, aggregated and normalized data

•Build user/ consumer app with security certification (e.g. PCI) instrumented to capture anonymized data

Business Model

•Clear case for a customer to pay

• Two modes: 1. Standalone 2. Bundled with Services

•Case#1: Define specific value proposition for clients to pay for

•Case#2: Ensure service bundle can be differentially priced to justify product development economics

Analytics Scope

•Specific and targeted solution for business problem; not generic service

•Not just descriptive

•Predictions, Recommendations & Outcomes

• Possibly for a area that is new, with poor client competency and requires capability augmentation

Differentiation & Competitive Advantage

• Does the data product help the existing business model differentiate well

• How robust is the differentiation

• What investment in the short and long term does this justify

Time Scale

• Have clear understanding of time scale for development and business integration

• How quickly- Can you reuse, repurpose, build and/or lease

• Use APIs [e.g. Watson] to quickly turn around problem statement

•Agile Development framework with scrums tracking on specific delivery

Management decision to build Data Product needs to be guided by the following principals

Decision of Engineering, cloud, Analytical platform and software are completely upstream decision9/18/2015 © Sumit Dhar 6

Toll Gates to graduate Data Product Ideas

• Assess Need/ Competition

• Opportunity & Biz ModelMarket Assessment

• Validate with friendly foes

• Outcomes & Biz ModelConcept Validation

• Identify Early Adaptor

• Beta Testing is key to successEarly Adaptor

• Share incremental results to start showing value and direction

Concept Validation

• Be agile and build a minimum viable productMVP Build

• Launch limited pilots with clear success criteria

• Syndicate winPilot

• Aggressively market and socialize win

• Align Sales & Marketing

Scale Up Wins

• Plan SaaS hand off to Analytics Service/ OperationsService Orientation

Data Product Development is not a back office endeavor; This requires considerable market integration and agile iteration based on market/ client asks

Appendix-B: Contains some “top of my head” use cases of ideas that I have come across and

9/18/2015 © Sumit Dhar 7

Data Product Enabled Analytics Delivery ModelData Product development and delivery with the objective to solve specific problem using cloud and API integration

Data Product

R&D

SaaS/ Business

Implementation

Analytics

Service

Business App

Prediction App(Scoring, A/B Test…)

Decision Levers(Interacts, Content, Offer)

Context Information(Social, Mobile, Device,

Action, Macro)

Discovery & ML (Prediction, Insights &

Recommendation)

Predictive Model(Intent, Action…)

Optimization & Outcome

Insights

Advisory

Focus on solving hard business problems can help create a niche 9/18/2015 © Sumit Dhar 8

The Role of a Big Data Products Leader

o The Leader is a market facing role who works with market leads to understand, enunciate and prioritize ideas for investment

o Has capability to set up a data product innovation playbook and put in place strong execution

o The Leader may not revenue target but may own overall investments targets and new marketable innovations target

o The incumbent needs to be have a decent handle on Data Science, Mobile/ Cloud Technologies but have a strong business orientation

o Fundamentally a strong people person who has the following competencies:

- Creates stakeholder engagement/ alignment internally to drive innovation

- Strong leader who can attract talent and build a strong team around him

o Be a go to person with respect to any conversation on Data Science internally and with client

o Build a best in class Engineering and Data Science organization and lead in a agile manner with scale

o Potentially be aligned to develop and manage a multi-cultural and multi-shore team as one organization

The Big Data Products Leader Plays a integrator role across the organization for driving analytics services paradigm shift

9/18/2015 © Sumit Dhar 9

Postscript…

o Bulk of the Big Data Product ideas are based on the “Data Exhaust”

- Data foot print created due to business activities

- Bread crumb data created due to customer actions

o Aspirationally building a product without strong business hypothesis can lead to long time cycle to failure

- Strong alignment of Data Science, Subject matter expertise and Services is required to drive success

o “Pivot or Persevere” It is necessary to quickly change direction based on market feedback or results to ensure optimal investment of man power

o The idea of Data Product based service delivery is transformational; requires significant organizational alignment

o The ability to develop a business model around such a concept also depends on the “buyers of services” and ability of market leaders in attuning clients/ prospects in the line of thought

o Ability to drive outcomes in a client business environment is dependent on following:

- Predictions impacting pertinent metrics

- Significant maturity in design of Engineering integration components

- Curious, open and data driven culture in developing client relationship to drive change

o Sponsorship and agreement in investment and go to market is required for development, socialization and marketing of Data Products

Building Data Product for fortifying a services business is a exciting proposition

9/18/2015 © Sumit Dhar 10

Sumit Dhar- Brief ProfileSumit has 18+ years of experience in the “Data Industry” with primary focus on Customer Insights & Analytics using Big Data. Sumit has worked and advised Fortune 500 companies in the US, UK, Western Europe and Asia Pacific in Financial Services, Healthcare and e-commerce domain. Currently he is with [24]7 iLabs and is responsible for driving outcomes and optimization using Data Science

Passionate About:- Data, Design for Analytics, Data Science

Interested in:- Building Business Models leveraging Data Science & Advanced Analytics

Knows Well: Data & People

Vice President- Client analytics & Data Science[24]7 iLabsEmail: [email protected]: +91-9686760383Twitter: @sxd213Linkedin: https://in.linkedin.com/in/sumitdhar71

9/18/2015 © Sumit Dhar 11

APPENDIX

APPENDIX-A: Organizational Alignment

• Sponsor client connects- decide on client we need to BU Owners

• Generate “next leap” ideas from specific clientsClient Relationship/ Account Management

• Validate problem statements and generalize across the market Market Leaders

• Infrastructure, cloud asset support and virtualization supportCTO Organization

• Own execution and competency framework- curiosity driven & evidence inspired work forceAnalytics Organization

• Structure more SaaS type contracts; Data and IP issuesLegal & Compliance

• Own Business case, investment time frame and Data Product scrumsData Science & Product Leader

• Develop MVPs/MVAs and evolve final work product; Own methodologyData Product & Data Science Org

9/18/2015 © Sumit Dhar 13