Positioning Your Analytics Team on the Right Projects
“Good soccer players follow the ball — the best soccer players are already there when the ball lands”
The project you’ll work on will have a large impact on the success of your team (and your own success). To some extent, this is one of the most important decisions you have to make every month / quarter / day (depending on your prioritization process).
Yet — I never really came across a mental model for the analytics world that fit my needs, to help decide which projects to work on. So this is the exercise we’ll be trying in the article — coming up with a mental model to help make better prioritization decisions.
There are risks associated with any project
As we discussed in the previous article, there are multiple macro-elements that can impact your study: Data Availability, Skillset, Timeframe, Organization Readiness, and Political Environment.
Each of those elements can generate some risks in your study. If you don’t have the right data, you most likely won’t be able to get the right answer. If you don’t have the right skillset, or if the distance between your current skillset and the one you need to acquire is too high, or if your organization is not ready to implement any of your results, or if there might be some political hurdles to the implementation — you have a higher risk of your project failing.
While there are a lot of different and complex forces at play, on your side, you can score your potential project in a simple way — using a 1–5 scoring system (or 0–100%), based on your confidence in your ability to overcome any of the hurdles we mentioned above.
Impact comes in different size and shape
Understanding how impactful a project is requires you to properly understand how your analytics project will be used in real life. To some extent, this is a good forcing function to properly understand the “job to be done” and how your work is going to be operationalized.
There are several things to consider while doing this exercise:
- It is important to consider “second order” impact. Some data projects will have an impact “by themselves”. Others will enable other people to have more impact (e.g. dashboard). Others will unlock some data analysis that were “locked” until then (e.g. data pipeline).
- The “depth” of the value is not the only criteria to take into account — but the “width” also. For instance, building a dashboard might generate a small amount of value but for a large number of users — which in aggregate is actually a lot of value.
- Impact needs to be considered with regard to the temporality of the work and the need. If your industry is rapidly evolving, or if there are some early signs of a shift in the company strategy — it is important to factor that in in your impact calculation.
Similarly here, using a 5 point scoring system can be an easy way to stack rank the different projects and understand which ones might generate the highest value.
Your time is limited
Time is another great forcing function here — because to accurately account for the time required to complete your study, you have to have a good understanding of the scope of the project. It is not just about the data project itself, it is about all those things that will make the project successful:
- Making sure everyone is aligned pre-project on the goal and deliverables
- Doing the actual project
- Making it digestible
- Communicating it to your audience
- Reaching alignment on the results and action items post-project
Note that there will always be this one curveball that comes flying at you halfway through the project. That’s the beauty of any data project, it is like Forest Gump’s box of chocolate — you never know what you’re gonna get. It is important to factor in these nice “surprises”, and it is always good practice to leave slightly more time for that. If you are in a situation where you are clueless about the time it will take, here are some options for you:
- Benchmark with similar projects done in the past
- Use the wisdom of the crowd and ask other peers how much time they think such a project would take.
- Run a “thought experiment” (but a real one, not something to which you think for 2 minutes). Take some time to visualize the full process, and all the steps you will have to go through to finalize the project. From your visualization exercise, assess the time it will take you from start to completion.
Putting it all together
Now you have a good understanding of:
- The time each project would take
- The impact they would have
- The risk associated with each one of them
You can put all that together and define your “investment thesis”.
Basically just like a VC fund that is investing into startups — you decide where to invest something that is way more valuable than money: your time. You can choose (within reason) how to balance the portfolio of your team: what is the risk level you are comfortable with? Do you want to go for a few moonshots or a multitude of easy / proven projects? Do you want to double down on projects that have already shown successful impact in the past? You decide.
My personal investment thesis: I always try to go for one or two low confidence / high reward projects per quarter along with a multitude of smaller / easier / with guaranteed value projects.
In short
Think of yourself as a venture capitalist in the world of data projects. Every choice you make is an investment — not of money, but of something even more valuable: your time and energy.
You’ve got all these potential projects vying for your attention, each with its own risks and rewards. Like a VC, you need to pick the ones that promise the best return on your investment. This means sometimes going for the long shots — those low confidence, high reward projects that could really pay off. Other times, it’s about stacking up those smaller wins, ensuring a steady flow of value and progress.
Ultimately you are in charge of defining your own portfolio — and it is important to take the lead on this activity, because if you don’t, someone else will.
