How to Conduct Customer Interviews for AI Products: A Founder's Guide

Let's be honest - interviewing users about AI products is tricky. Your users probably don't know what's possible with AI, might be skeptical about it, or worse, think it's magic that can solve everything. I've been there, and I'm going to help you navigate these waters.

Why AI Product Interviews Are Different

When I first started conducting interviews for AI products, I made the classic mistake: jumping straight into features and capabilities. Users would either get overwhelmed or start imagining unrealistic scenarios. What I learned is that you need to start with their current reality, not the AI.

The key isn't to talk about AI at all - at least not at first. Instead, let's focus on what actually matters: the problems your users are trying to solve.

The B2B vs B2C Reality Check

If you're building a B2B AI product, you're not just dealing with users - you're dealing with entire organizations. The VP of Engineering who loves your AI capability might not be the one making the purchase decision. The end-users who would benefit from your tool might not be the ones evaluating it.

Here's what actually works: Start by understanding the organizational workflow. Who touches the problem you're solving? What happens before and after? This context is gold. One founder I worked with discovered that while their AI document analysis tool was impressive, what really sold it was how it fit into existing approval workflows.

For B2C AI products, the game is entirely different. Individual users don't care about enterprise integration or ROI calculations. They care about immediate results and personal benefit. Think about face filter apps - users don't care about the sophisticated AI running in the background. They care about looking good in their photos.

The Jobs to Be Done Approach That Actually Works

Forget the theoretical frameworks for a minute. Jobs to Be Done (JTBD) is really about one thing: understanding what drives people to look for a better solution. For AI products, this is crucial because people often don't know they need AI - they just know their current solution is frustrating.

Instead of asking "Would you use an AI tool for X?", try this conversation starter: "Tell me about the last time you got frustrated with [current process]." Then dig deeper: "What did you do about it?" "Why was that particularly frustrating?" "What did you try next?"

I was recently helping out a founder researching customers for an AI email management tool. The breakthrough came when we stopped asking about email features and started asking about their worst email days. The stories poured out, and the real opportunities became clear.

Getting Real About AI Expectations

Here's a truth many founders miss: users' concerns about AI are usually more revealing than their feature requests. When someone says "I wouldn't trust AI with that decision," they're actually telling you something valuable about their decision-making process.

Instead of defending the AI, dig into their current trust indicators. What makes them trust their current solutions? What checks do they perform? This is how you learn what your AI needs to explain, display, or validate to earn trust.

The Implementation Gap

For B2B AI products, there's often a massive gap between "interested" and "implemented." One enterprise AI tool I worked with had enthusiastic buy-in from every demo but struggled with actual adoption. Through customer interviews, we discovered that the real barrier wasn't the AI - it was the manual work needed to get started.

This is why you need to interview not just decision-makers, but also the people who would handle implementation. They'll tell you about the practical barriers your sales team will never hear about.

Making Interviews Scalable

As your AI product evolves, you need a consistent way to gather user insights. This is particularly crucial for AI products because user expectations and understanding of AI capabilities change rapidly. Using tools like Resonant, you can automate these interviews while maintaining quality and consistency.

The key is to keep gathering insights even as you scale. User perceptions of AI are evolving rapidly, and what was true six months ago might not be true today.

Real Talk About ROI

For B2B AI products, everyone will ask about ROI, but here's the catch - they're often asking the wrong question. Instead of focusing on pure cost savings, dig into the ripple effects. One AI scheduling tool found that their biggest value wasn't in time saved scheduling meetings - it was in reducing the mental load of their users throughout the day.

The Path Forward

Start your next user interview by parking the AI discussion at the door. Instead, focus on understanding their current world deeply. The opportunities for AI will become obvious, and more importantly, you'll understand how to position and build your product in a way that actually resonates.