- DS is not a doc that takes months to develop and is kept aside to be given to new recruiters when they join. It needs to be revised and updated as the business evolves.
- It should contain both the engineering and infrastructure approach, as well as how you plan to structure the organization and teams to make the most of your data.
- It needs to be well integrated with the rest of the business and should possibly be the north star with all other plans emerging or linking back from it. It should NOT exist in isolation.
So how do you build a data strategy?
Now, this depends on what stage of the customer acquisition journey you are on. Eg: a newly launched startup would like to acquire users rapidly while keeping in mind retention and engagements are high. Here, the data team must focus on building the right foundation to collect and analyze the necessary data to make rapid iterations on the product if necessary.
Next would be to identify your data use cases so that one can focus on the ‘Data Requirements’, ‘Key Project Resources’, and ‘Technology’. It is useful to think of these areas for all data use cases together, so you can spot overlapping requirements across projects and better coordinate resources.
We should also aim to look at ‘Data Architecture’ and ‘Data Governance’. Data Architecture is important because as a business scale so does the amount of data. At an early stage, firms can still switch from one architecture to another but it may not be feasible after a while. Eg: Slack spent 3 years (from 2017 to 2020) migrating their data architecture from active-active clusters to a tool called Vitess. Hence, be mindful of your data requirements and be agile and keep iterating towards product-market fit.
Lastly, do keep a lookout of data governance as they are required by law under regulations such as the General Data Protection Regulation (GDPR) or the California Consumer Privacy Act (CCPA).
So as you build your organization, do remember that data collection can be expensive in terms of time and money. A well-thought-out data strategy helps to zoom in on which data will unlock the most value and hopefully kick off the data flywheel.
Register now to get access to the full AI Case Studies Bible, and the Decision Maker’s Handbook to Data Science!