“Move fast, break things” isn’t new, but AI makes this motto even more palatable. AI and datafication are great enablers for rapid decision-making, it’s true. So, is Research a thing from the past? From when we didn’t allow ourselves to make mistakes, when we could now look at a product’s performance as usage happens? Now that the mindset is “ship at 70% readiness” (which is the equivalent of “we’ll fix it in post”), what is left for Research?
It seems that qualitative attitudinal research, which UX Researchers do a lot, is in a tough spot. CX Researchers seem to still have an ear because of the quantitative part, but they’re still focused on attitude (CSAT, NPS). Moreover, I think it’s obvious to say now: companies don’t really care about their users. They care that numbers are green and lines go up. Happy customers are one lever among others. If you drink the kool-aid and genuinely believe your product is the best, so much that you’ll crush the competition, then why would you care about someone who might make you think twice about your belief? This behavior is solid ground for enshittification: you optimize for metrics long enough, and the metrics start lying to you. Engagement climbs because the product is coercive, not because it’s good. Retention holds because switching costs are high, not because users are happy. You’ve replaced the signal with its shadow.
However, we haven’t seen a clean, large-scale reckoning yet. There’s no obvious cautionary tale you can point to and say “that’s what happens when you ignore your users”. That absence is actually the problem: it means the feedback loop is long, and companies keep mistaking silence for validation.
So, is UX Research fucked?
There are two edges where UXR still matters, and they share something: being wrong costs something real. One is high-stakes, embodied contexts like physical environments, regulated industries, anything where an app can’t capture what actually happens. There’s documented evidence of people getting burned by induction stoves with touch-only controls because the interaction model was designed for screens, not kitchens. No engagement metric catches that. Field research does. In a sense, this is where the roots of UX Research are: Human Factors.
The other is trust-dependent products: smaller, niche tools where users have real alternatives and the willingness to leave. The loyalty is fragile and earned. When Firefox started nudging AI features on users, the pushback was immediate and the team walked it back. That’s not the norm, but it’s a signal that some products still feel the cost of losing the room.
Research is being filtered back to seriousness. It’s dissolving from the bloated center, from the orgs that hired twenty researchers because it looked mature, and it’s surviving at the edges where getting it wrong has consequences. That’s not the worst outcome. It just means the job is harder to get and more important when you have it.