What Makes A Successful Data Team?
Introduction
The #1 thing that all data team should strive for is to be: Value-Driven
It might seem vague at first but it is an important mentality to embed in your team: what we do serves a purpose. Every data project needs to drive value for the business and its success should be evaluated in this manner. Not how sophisticated the algorithm is but how it makes an impact on your products, services, customers and profitability in quantifiable ways. A sophisticated algorithm is important but it is not the end goal otherwise you risk over-engineering the solution and a diminishing return on your data investment.
How To Ensure And Measure Value
First you need to align the data strategy with your business model and growth objectives. If you have a subscription model and are racing towards $100m ARR, your KPIs should contain a list of contributing factors across all teams that collectively get you to this target.
You might decide to create predictive models to flag users at risk of cancelling the subscription. The first question should be how this model will be deployed by the business to generate ROI? Which teams should I involve and what tool stack do they currently use?
You also need to be specific in evaluating the success (or failure). For example, you embedded a propensity scoring model into an automated churn prevention CRM, which led to a 10% reduction in membership cancellation. Based on the current size of your subscription base and pricing, it is estimated to be a $5m addition to your ARR. Be specific and objective in terms of the outcomes. It helps you prioritise resources and get future buy-in from key stakeholders.
“Helping us make better decisions” is a slogan, not a good way to evaluate and progress your data team.
Speed To Value
Once you define the data strategy, the other element of value to ensure is the speed to value. How long does it take for the business to benefit from your current initiatives? There is an opportunity cost when you don’t get to incremental value quickly. Opportunities cost both for the business to action on insights and for the data team to demonstrate impact.
To get to incremental value quickly, you need to create a clear data roadmap, prioritise the low-hanging fruits and iterate from there. Break the project down into smaller milestones and deliverables.
Don’t try to do everything at once. For example, instead of a 6-month marketing analytics project, we recommend breaking it down into smaller pieces: first 4 weeks to centralise marketing reporting across Facebook and Google, then 4 weeks to model first-party data for campaign performances, followed by attribution analysis and automation, probabilistic models on non-consented users, LTV predictions for early ROI signals and more. Each step there is concrete deliverables for the business to start making use of the insights generated.
Data Adoption
There is little value from a data solution that is not being used by your teams. With every delivery, plan a session with business users to take them through the solution and show key use cases. This could be a one-on-one meeting with the product manager to go through the findings from an in-depth analysis or a structured training session with the whole marketing team.
Communicate Data ROI
We talked about measuring the value of your data initiatives. It is also important to quantify the investment in data, in terms of staffing cost and infrastructure investments. With two sides of the coin, you will have a good understanding of your data ROI. This will help you prioritise resources in the company and identify future high-ROI data projects. Finally, make sure to regularly communicate data wins and successful use cases across the organisation to increase data awareness.
Conclusion
We cannot emphasise enough the one and only goal of any data teams: to generate value for the business. This ethos will guide your team, shape its culture and lead to success for the team and the company.
Be specific and objective when evaluating data projects, plan incremental milestones to get to value quickly, onboard business users to the data solution, and measure and communicate data ROI from the moment you see results.
A Checklist To High-ROI Data Projects
- Align data strategy with business objectives. Involve all stakeholders at this stage.
- Establish business priorities and potential returns. Start with a list of business challenges and questions. Put a value on solving these items.
- Build a data roadmap to address each business requirement with smaller milestones and allocate resources accordingly. Don’t try to tackle everything at once.
- Implement and monitor progress over each milestone.
- Onboard business users to the data solution with release and training sessions. Gather feedback and improve on usability.
- Measure ROI and communicate data success stories as soon as results are available.
- Bonus point: stop optimising the 1%. If solving a problem that will only increase gains within 1% of your customer base, the problem is probably a very low priority one. You need to work on moving the needle where there is significant impact.