Ethnography finds what interviews miss. Almost no B2B product team does it.

How to find hidden needs, and three tried & tested ways to get yourself in the room

I ended my last piece with a prediction: AI is outpacing humans at data analysis, but humans keep the edge on data collection.

No LLM - however good it gets at reading transcripts, synthesizing Gong calls, or picking out customer opportunities - can sit in your customer’s office and observe how they actually work. The awkward workarounds. The steps nobody documented. The thing they do every morning that they’ve completely stopped noticing.

That’s what ethnographic research gives you.

And almost no B2B product team does it.

Contents

• Your discovery toolkit has a blind spot

• Hidden needs

• Observe like an anthropologist

• Getting in the room: three ways that actually work

Your discovery toolkit has a blind spot

Ask most product teams how they learn about customers, and you’ll get some version of the same list: customer interviews, sales and CS call analysis, A/B tests.

These aren’t bad methods. They’re just incomplete in a very specific way.

• Analyzing sales and CS calls tells you which feature requests, questions, and complaints come up most. Useful signal. But you’re in reactive mode - reading a list of things customers have already decided to bring up, in a context where they’re already frustrated enough to contact support. If you don’t understand the customer opportunity beneath the feature request (”what problem are they actually trying to solve”), you’re at very high risk of implementing the first, but not the best solution idea, or solving a problem that isn’t really that important.

• Story-based interviews solve that. You dig into customer pain points, wishes or desires that have actually come up in the past. Done well - à la Teresa Torres - you get a real account of a recent experience: what the customer was trying to do, what got in the way, how it felt. But you only capture what they bring up, what’s on their mind right now. You miss their environment, their workarounds, what a Tuesday morning actually looks like, their hidden needs.

  
  An adjacent problem - especially for product teams serving niche B2B audiences - is that product managers struggle to get beyond a superficial understanding of the customer problem they’re solving. Even a great 60-minute interview barely scratches the surface of what their domain actually looks like. It’s like trying to build a deep friendship over monthly calls. Spending a weekend together lets you go much deeper.
  

• A/B tests tell you what users do (which is obviously more reliable than what they say they will do). Downside: Zero insight on why; and you only see how they behave in your product surface.

Interviews and call analysis surface what customers can put into words. A/B tests show you what they do inside your product - but not the larger context they’re operating in, and not why.

None of these reach the need the customer hasn’t consciously registered.

Hidden needs

The Kano model describes three types of customer need:

• Performance needs - what people articulate in interviews. “The reporting is slow.” “I need better export options.” These are the needs that get requested, tracked, and built.

• Basic needs - what customers take for granted. They don’t mention them because they assume you already know. When they’re missing, customers are angry. When they’re there, invisible.

• Hidden needs - needs customers haven’t consciously identified. They experience the problem, but they’ve normalized it or never named it. Solving these is what produces breakthrough products, not incremental ones.

(Graph taken from “Identifying hidden needs: Creating breakthrough products”, by Keith Goffin and Ursula Kroners)

Sequoia’s PMF framework adds a fourth category worth naming separately: hard-fact-of-life needs. The customer is aware of the problem but has mentally filed it under “unsolvable.” They’ve resigned to it - so they don’t bring it up in interviews, not because they don’t feel it, but because they’ve stopped expecting anyone to fix it.

Before 2022, I had to pay a development agency around €100k to get my app-idea development, and I accepted that as a hard fact of life. Look how that changed.

Understanding hidden and hard-fact-of-life needs is key to building an innovative, differentiated product.

If you solve the problem that the customer tells you about, you’ll hopefully meet their expectation.

If you solve the problem they weren’t even aware they were having, or they considered unsolvable, you exceed it.

If you solve the problem customers are talking about, you’re facing stiff competition. Obvious pain points have 100s of software tools. Almost every ocean is a blood-red ocean. And some team are n8n’ing and Lovable-ing their way to their own “ideal” tools.

According to research done in 2006, only 15% of organizations use ethnography - despite it being rated the most useful research method. Especially if you're building for a niche B2B audience whose real-world context is hard to grasp from a few 60-minute calls, ethnography is the gap your competitors aren't filling.