How To Pull Stories From Data and Make Them Actionable - Issue 251
Finding the Why: turning raw data into clear, actionable insights and stories
Last week, I ran a workshop on making data more actionable and wanted to share some of the key takeaways here as well.
I often get questions about how to uncover the “why” behind a trend, how to present findings in a digestible way, and how to turn large volumes of raw data into insights that drive decisions. So, I decided to consolidate some of the techniques I rely on or adopted from other great analysts that help transform data into (hopefully) clear and actionable insights.
So, read below my methods and tips on how to find a story or pattern in the data mess, what to do if you can’t find one, and how to present it in a way that is actionable.
1. It’s not actionable if no one will use it.
This might sound obvious, but it doesn’t stop us from spending hours, or even days, on reports that ultimately go nowhere.
Before diving into techniques, it’s important to understand that no data insight is actionable if there is no execution layer to support it.
Sure, it’s useful to know that Brazil contributes 20% of your total DAU. But unless this insight is communicated to the marketing manager, who can target that audience with a campaign, or the product manager working on localization for a new feature, it won’t be actionable.
Before you begin any analysis, you should clearly understand who the recipient is, what domain they own, and why they asked you for this report in the first place.
Insights don’t mean much if they’re not executed (although doing nothing is often wise). The report or dashboard you deliver should be integrated with operational workflows. It should be connected to a feature owner, stakeholder, or team with the authority to act.
Often, it’s on you to make sure the report doesn’t just exist - it’s actually seen and used by the people who can do something with it. When I say “you”, I mean your manager or whoever is leading the analytics domain at your company.
2. The “So What?” Moment
Analytics is challenging because people often have different expectations from the same data point. Analysts want to take credit for enabling and presenting data stats only to be met with the dreaded “So what does this mean?” Or worse: “What am I supposed to do with this now?”
What elevates your analysis is asking, “So what?” 5 to 7 times for every data point you report. Here is an example:
The payment success rate is 25%. → So what?
This means the payment failure rate is 75%. → So what?
About 96% of these failed payments are from international transactions, primarily in Asia, India, and the Philippines. → So what?
Stripe doesn’t accept payments from these countries. → So what?
We need to pause marketing initiatives in these regions until we can process international currencies. → Action!
Stakeholders (especially executives) generally expect only the final insight in #5. They don’t have time (or patience) to wade through every layer of reasoning. However, if not trained, most analysts stop at #1, assuming the job is done. If there’s no senior partner to connect the dots, the insights quickly get buried under a pile of half-finished reports. It’s the analyst's responsibility to get to #5 within the initial report, not the stakeholders’ job to follow up questions and requests.
It takes a long time to get from #1 to #5. You need domain knowledge, historical context, stakeholder input, lots of data checks—and even then, there’s no guarantee you’ll land on something actionable.
Not every data point you report will have an answer to every "So what?" question, and that’s okay. What matters is training your analytical thinking and developing the habit of looking beyond the surface to see the bigger picture.
