I’m going to tell you something that might sound crazy. Most of what you’ve read about AI user research is wrong. Dead wrong.
I’ve seen it from the inside. I’ve built and sold two companies, invested in over 200 more, including some of the biggest names in AI like Anthropic and OpenAI, and I’ve seen the same pattern again and again. Founders are drowning in data but starving for insights. They’re using so-called “AI-powered” tools that promise the world but deliver a pile of neatly organized, utterly useless reports.
Why? Because these tools and the guides that promote them are designed for big, slow, bureaucratic companies. They’re not for founders who need to move fast, break things, and actually talk to their users. They’re a cargo cult of user research, a performance of understanding without any of the real substance.
I remember one of my early startups, MovieLaLa. We were trying to build a recommendation engine for movies. We had all the data in the world. We had user ratings, watch history, you name it. We even had a fancy algorithm that could predict what you’d want to watch next with 90% accuracy. And you know what? Nobody cared. The product was a ghost town.
We were so focused on the “what” – what people were watching – that we completely missed the “why.” Why were they choosing that movie? What was their mood? Who were they with? What were they hoping to feel? The data couldn’t tell us that. Only real, messy, human conversations could.
That’s when I learned the hard way that user research isn’t about data. It’s about empathy. It’s about getting inside your user’s head and seeing the world through their eyes. And no AI, no matter how smart, can do that for you. At least not in the way most people are trying to use it.
The Amazon Way (and Why It Fails for Startups)
I’ve had a chance to see how the big guys do it. I’ve spent time inside some of Amazon’s most secretive labs, the places where they cook up the future. And let me tell you, their approach to user research is… thorough. They have teams of PhDs, massive data sets, and a process for everything. They can tell you with statistical certainty that changing the color of a button from blue to slightly-less-blue will increase conversions by 0.01%.
And for them, that’s great. When you’re operating at Amazon’s scale, a 0.01% improvement is worth millions. But for a startup? It’s a rounding error. It’s a distraction. It’s a waste of time and money that you don’t have.
Startups don’t need statistical certainty. They need direction. They need to know if they’re building something people actually want. And for that, you don’t need a team of PhDs. You need to talk to five users.
Yes, five. That’s it. Just five real, live human beings who have the problem you’re trying to solve. Talk to them. Watch them use your product. Ask them open-ended questions. And then, most importantly, shut up and listen.
The Counterintuitive Guide to AI User Research for Founders
So, how do you do user research the right way? The fast way? The way that actually gets you results?
Here’s my counterintuitive, step-by-step guide for founders who need to move fast and get results without a massive data science team.
1. Forget the AI (for now)
I know, I know. This is an article about AI user research. But the first step is to forget about the AI. Seriously. Put it out of your mind. Don’t even think about it.
Your first and only goal is to talk to your users. That’s it. Everything else is a distraction.
2. Find Your Five
Who are your five users? They’re the people who are so desperate for a solution to their problem that they’ve hacked together their own. They’re the ones who are using a spreadsheet to do something that your product could do in a click. They’re the ones who are complaining on Twitter about how much their current solution sucks.
Go find them. They’re out there. And they’re waiting for you.
3. Have a Real Conversation
This isn’t a survey. This isn’t a focus group. This is a conversation. A real, human conversation.
Ask them about their life. Ask them about their problems. Ask them about their hopes and dreams. And then, and only then, ask them about your product.
And when you do, don’t ask them if they like it. Ask them what they would change. Ask them what they would pay for. Ask them what would make them tell all their friends about it.
4. Now, Bring in the AI
Okay, now you can bring in the AI. But not in the way you think.
You’re not going to use the AI to analyze your data. You’re not going to use it to generate a report. You’re going to use it to do the one thing that it’s actually good at: transcription.
That’s it. Just use the AI to transcribe your interviews. That’s all you need it for.
Why? Because the act of listening to your interviews again, of reading the transcripts, of pulling out the key quotes and insights – that’s where the magic happens. That’s where you start to see the patterns. That’s where you start to build empathy.
Don’t outsource that to an AI. That’s your job as a founder.
The One Metric That Matters
So, how do you know if you’re on the right track? How do you know if you’re building something people actually want?
There’s only one metric that matters: the “shut up and take my money” metric.
If you’re not hearing that from your users, you’re not there yet. Keep talking to them. Keep iterating. Keep building.
And whatever you do, don’t get distracted by the siren song of AI-powered user research. It’s a trap. A beautiful, data-driven, statistically significant trap.
Your users are out there. Go talk to them.
Frequently Asked Questions
Can these results be replicated?
The specific numbers will vary, but the underlying patterns and principles are transferable. The key is understanding the context behind the results, not just copying the tactics. Every company has unique constraints that shape what works.
What would you do differently looking back?
I'd move faster on the things that were working and cut the things that weren't sooner. Most founders, myself included, hold onto failing strategies too long because of sunk cost. Speed of learning is everything.
What was the biggest challenge in this case?
Almost always, the biggest challenge is people and alignment, not technology or strategy. Getting the right team focused on the right problem is harder than any technical challenge I've encountered.
How long did it take to see results?
Most meaningful business results take 3-6 months to materialize. Anyone promising overnight success is selling something. The companies in my portfolio that grew fastest were the ones that stayed patient and consistent.