Data Strategy - By Avinash Narasimha

Vision & Sponsorship

• Why do we exist, what purpose do we serve, how do we create value? - Serve business priorities, treat data as an asset and create a competitive edge for the business, adopt data centric approach, enable Digital Transformation, enrich quality insights from AI, solve existing pain points

• Sponsorship - Having clear sponsorship from the business is key to success; alignment from business before starting is essential, I have heard of data products being built with no users which should be avoided at all costs. It is important to challenge business to define their long term priorities which could help define a better Data Strategy. If priorities are already defined, then its good to follow an agile development cycle with clear milestones, adopt fail fast approach

Data Governance

Establish policies and procedures for data management, including data quality, security, compliance, and privacy. This ensures accountability and proper handling of data assets (stewardship, policies etc)

Operating models

• How do we engage with the business?

• Do we have alignment at the right levels?

• Do we have clear sponsorship with business leaders?

• How do we operate on a day to day basis?

• What kind of governance is needed?

• How do we track spend & show value to the business?

Tools/Tech

• Have a clear data product strategy – definition, why is it needed, adoption, maintenance, cost, user personas and how they will consume the data

• Define platform strategy – buy vs build, maintenance, cost, long term vision

• Modern Data Stack - Implement a modern data stack that supports data ingestion, storage, processing, and analytics. Choose tools that are scalable and fit the team’s needs.

• How does the next gen tools impact what we do today? (Ex- delta lakes, Fabric, knowledge graphs, DataMesh, GenAI etc)

• How do we enable AI teams to go faster? It could be through self service models, API’s, datacubes, data products or even Databricks/Snowflakes etc

• What kind of architecture patterns/reusable components must be built which delivers expedited value for the business?

Talent

• What kind of talent is needed to execute the strategies?

• How do we structure the Data teams?

• What roles are needed to help Data teams be successful?

• What skills are needed in the next 2-3-5 years?

Some other points to keep in mind

• ROI from Data teams - Measure of success- Adoption, quality of insights, FINOPS, revenue impact ($), cost savings, time savings, self-service models, enhanced customer experience, feedback from customers; Data teams have historically struggled to show value, hence the frequent churn at the Chief Data Officer levels

• Demand vs supply – Focus more on the demand than the supply; Most of the time spent should be with business to cater to their requirements than figuring out what could be useful; wherever possible have someone from the business be the Product Owner (skin in the game)

• FINOPS needs to be adopted from day 1, it should become a cultural thing for the teams rather than doing an exercise to review and lower costs once every 3-6 months.