Data Strategy: Where to Start and How to Do It Well

Irina Goloshchapova25 May 2024

Problem

A new head of analytics or Chief Data Officer (CDO) must decide what the data office does over the next two to three years. The scope is wide: data collection, quality, warehouses, analytics, ML, people and governance. Without a frame, the plan shrinks to tasks for the next two months, or to projects done for their own sake, like a data platform nobody needs. The cost is lost direction and no way to show progress.

Example from the article: the Operational Block of Raiffeisenbank has more than 2000 processes to optimize and many systems producing data of criticality level 0. The speaker built the first strategy of its Data Office and analytics with the frame below.

Solution in brief

Start from three inputs: a maturity assessment, stakeholder interviews and the company and IT strategies. Turn them into vision, value and metrics, then work through four sections: people, operating model, data governance, technology. Set target levels only where the company needs to go, and check the result against ten signs.

Questions this case answers
  • How to build a multi-year data strategy for a data office from scratch using a maturity assessment, stakeholder interviews and the company strategy
  • How to check that a data strategy is good and will be adopted, using a metrics map along the data product chain
  • How to build a data strategy for a data and analytics office from scratch
  • How to assess data management maturity and choose which gaps to close first

Tools and methods

  • Data Strategy Frame — three inputs, vision/value/metrics, four sections
  • Data Management Maturity (DMM) assessment — scores about 12 data management sections plus two analytics sections, current versus target level
  • Data Office metric map — metrics along the path from sources to the business user
  • Star map and FTEM — competency gaps in the team
Continue reading — it's free

Google or a one-time email code. No password.