Why does being data-driven so often mean forgetting the user?

And what can we do to fix it?

As Generative AI becomes more and more involved in the lives of product teams I think we need to take a moment to remember what it is we really do. After years of creating digital products to simply consume media (YouTube, Instagram, TikTok) and allowing those companies to decide the metrics for success and then mold their products to maximize for those metrics with negative results culturally (the nebulous “engagement” leading to rage baiting) we’re beginning to see some pushback. At the end of 2024 we saw reports that kids under 15 were spending less time online year over year (https://kidscreen.com/2024/10/28/feature-kids-are-actually-spending-less-time-on-devices/).  I believe that we are at a cultural inflection point. I believe we will begin seeing more products focused solely on solving business problems and fewer media consumption social platforms, a move that will do nothing but good. Unfortunately, as we see increased investment in those kinds of platforms I believe we’ll see increased usage of GenAI tools to design said platforms.

The Problem of the Data-Driven Product

The products we produce should help our users in their daily lives. To accomplish that goal takes user empathy and observation, two skills that are invaluable when it comes to product roadmapping and essential when plotting UX changes. In the 2010’s we started seeing new tools for capturing user patterns without communication or observation. Every startup wanted to tout its “data-driven design paradigm” and the various heatmaps and eye tracking software it employed to arrive at that data. That drive for data never really lessened and with it came a new frustration. As teams relied on automated feedback they would solve for the expression of the problem, but not for the problem itself. They were missing the color that would be added via conversations with users.

Let the people speak. Then listen.

Conversations are hard and occasionally awkward, I get it. But direct feedback from your end users is the lens you need when implementing feedback from automated data. Otherwise you’re building a solution for you, and not them. Being successful in a product role is just as much knowing that you’re not necessarily the final authority on anything. That your job is to solve problems for your users, not solve your user’s problems for yourself. If a particular demographic isn’t finding their way through your onboarding flow it’s important to recognize that you should talk to them to figure out a solution, not ideate on your own.

It seems simple, but it’s so important and I’ve seen so many brilliant product people say they don’t know why their product isn’t stickier, they’re making all data-driven product decisions. But when asked the last time they got on a call with a user or sent out a feedback survey they stare at their shoes. 

We’re building products for people. Let’s not forget them.