How to Choose Which Data Projects to Work On

As a business / data analyst / scientist, there is often plethora of projects to work on. Knowing how to select the right projects can have a tremendous impact on your career and the company you work at.

This becomes especially important when you start leading a team, but is not something that is usually taught. In this article, I am offering a simple methodology to stack rank different projects and decide which ones to invest time on.

#1: Understanding the realities of the data world

  • A lot of projects don’t turn into real-life impact. A few years ago, Gartner estimated that 60% of big data projects fail. The number was later deemed too conservative, with the true value being closer to 85% [1]. Inversely, a few projects can generate outsized impact.
  • Selecting a data project (and investing time in it) is quite similar to investing money in a startup — you can have a few ‘winner projects’ and a lot of projects that won’t make it to market. Using this perspective can help build your own time-investment thesis, and make sure you optimize the value you generate.

The time investment matrix for data project (image by author)

#2: Scoring your potential projects

  • Probability of success: How likely will this project be a success? And by success I mean — will have a real life impact, i.e. will trigger change and these changes can be attributed back to your project?
  • Impact: How impactful is this project going to be? How much money could your project generate for the company?
  • Time: How long will it take you to complete the project? The more time you spend on one project, the less time you have for others.
  • (Optional) Learning: What will you learn by doing this project? In case of tie, it can be interesting to factor in ‘long-term investment’ — basically how much this project will teach you/the company.

#3: Defining your time-investment thesis and your diversification strategy

  • Amount of time: How many projects can you take over a quarter? Based on your experience in the company, or if you are managing a team, this can change radically.
  • Risk appetite: How comfortable are you with risk? Do you prefer to play it safe, or can you take more projects that are inherently riskier?
  • How can you maximize your impact based on the above? Will you go with a few high reward, high risk projects, or would you prefer to have a lot of smaller projects with guaranteed return?

Example with a real-life scenario

Let’s imagine you are working for a Customer Relationship Management (CRM) company, and you are supporting the business team with their data needs. You start Q3 with the following five asks:

  • #1: The dashboard ask: The sales team would like a dashboard to track the sales of a specific product. The ask is pretty straightforward, but as this is a dashboard that will be used by multiple stakeholders, it will be time-consuming to reach alignment.
  • #2: The sales pitch ask: The partner management team is putting together a sales pitch to convert an important customer to a higher price point and would like your help. The ask is also quite straightforward and would only take a few days. The impact of this task is not necessarily important, but you believe there might be some potential if you find a way to actually scale the ask.
  • #3: The ranking ask: The acquisition team would like to get smarter with qualifying their leads and would love your input to rank the leads they have based on their likelihood of converting to paid customers. The model they are currently using is pretty simple and you believe there is a lot that can be done. The impact of such a project could be phenomenal, but building a model and getting internal buy-in will require a lot of work.
  • #4: The A/B testing ask: The sales team (again) is trying a new sales approach and would like help setting up an experiment to understand the potential impact. From what you can gather, it is not really a disruptive change — if there is any impact it should be quite incremental — and from a quick power analysis you did, there is a high chance that the experiment doesn’t return anything statistically significant. But the team still wants to go through with it.
  • #5: The “new metric” ask: .Your management would like to develop a new “activity churn” metric, i.e. a metric that tracks when a customer stops using the product, so that the partner management team can reach out to these “at-risk” customers as soon as possible. The project seems interesting, but a quick internal literature review shows you that the team has already been trying to set up such a metric for the past few years, and have failed each time (mostly due to internal politics).

Based on the context you managed to gather from all the different projects and your assessment of your own competencies, you score each project as follows:

An example of a project scorecard (image by author)

Which allows you to build the following graph, with x= Time commitment and y= Expected value:

Time commitment x expected value matrix (image by author)

From there, it seems that you should prioritize 3 projects: the Ranking project, the Sales Pitch project and the dashboard project.

Now, if you have time to do all those projects, all good! But if not, this is where you “time-investment” thesis comes into play — it will help you decide among those ‘ties’ and come with a final list. Now you are all set for the quarter!

Note: all the numbers in the table above should be based on your own understanding of the situation, your competencies, and same for your investment thesis. The table can vary widely for two different people with the same projects — this framework is simply a way for you to rationalize your decision making process.

Other tips to improve your decision making process

  • Look for project with cross-functional alignment. Ultimately the success of your project is dependent on if it will be implemented / used (which is partially not under your control). It is important to get alignment with the different stakeholders as early as possible, to make sure it will be a success.
  • Look for projects with the right timing. Good soccer players follow the ball — the best soccer players are already there when the ball lands. Understand where your industry is heading and be there.
  • Go deeper on winners. Sometimes it is easier and more impactful to double down on successful projects than to start another one from scratch or to try to fix a failing one. Be ruthless in your prioritization.
  • Review your investment strategy at the end of each quarter. Understand what you did well (or did not do well) and keep improving your strategy.


Original post:

Leave a Reply

Your email address will not be published. Required fields are marked *