Last Sunday, we were prepping for my eldest daughter's school application video—you know, the one where they ask her to introduce herself, talk about her interests, that sort of thing. In theory, simple. In practice, we all know how this goes with a four-year-old.

So I thought of a rough structure, explained it a little to her, and then we said, "Tell them what you like." And off she went.

Except somewhere in that sentence, "things I like," it morphed into a subtle statement—with a smile—about why she doesn't like books.

Not a scripted performance. Genuine kid frustration leaking out. But so subtly you had to decipher whether she was being cheeky (she does that sometimes) or serious about it.

Now, I could have (and have in the past) said, "That's not what we wanted, let's try again." But I found myself doing what I tend to do in these moments—when I pay attention: I stopped and listened.

Because I've had enough experience to know something: in life, as with building products or services, the edge cases are where the real story lives. Not in what people say directly. In what leaks out when they're not performing.

What I Learned by Actually Listening

The conversation that followed revealed something entirely different from what her initial position suggested. She wasn't rejecting reading. She was rejecting the experience of reading because of the frustration she felt in that specific moment—she was tired, she had something else on her mind, and every small failure to read properly chipped away at her confidence.

But then—and this is the part that matters—when I dug deeper, the real insight emerged. She wasn't excited about reading itself. She was excited about the potential of reading because it led to a few things: 1) Being classified as a "Big Girl", 2) Her ability to play big sister to her brother, and 3) Getting into the school she wanted. So while the surface objective was "learn to read books," the actual motivation was entirely different—and so the approach and incentive structure had to change.

One kid. One rant. One conversation. But in that single data point, I saw something that no survey would have captured:

  • A rejection of the current experience (not the category itself)

  • An unintended use case (reading to her brother, not reading for herself)

  • The real blockers (phonics foundation for technical execution, AND context—she needed to see reading as connected to her identity as a "big girl," not as a standalone task)

  • The actual lever that would unlock engagement (identity—becoming a "big girl," which meant being able to read to her brother AND getting into the school she wanted)

This is an edge case. Not a bug. Not an outlier to be dismissed. A signal.

And it's exactly what most product teams miss when they're drowning in aggregated data.

Why Edge Cases Matter

As a product builder, I've spent years optimizing for the center of the curve. "Our data shows 73% of users do X, so let's build for X." And that's not wrong—but it's incomplete.

The edge case—the unusual behavior, the unexpected objection, the subtle rant when someone should be smiling—these are your canaries. They tell you where the puck is actually traveling, not where it is today.

Customers don't always tell you directly what they want. They tell you indirectly—through what they reject, what they bypass, what they complain about obliquely. A 4-year-old saying "I don't like books" with a smile is communicating something real. But if you're not listening for subtlety, if you're only capturing the surface objection, you miss it entirely.

Edge cases reveal:

  • Why people reject your current experience (not just that they do)

  • Unintended use cases that could be new product opportunities

  • Foundational gaps that block mainstream adoption

  • Shifts in behavior before they appear in your metrics

The problem is that most product discovery methods are designed to be efficient, not revelatory. Surveys ask what you want to know. User interviews are scheduled and structured. Focus groups are moderated. They're all tools of confirmation, not discovery. And individual anecdotes—a single parent's subtle rant about their kid's reading habits—get lost in the noise of "N=1, not statistically significant."

Except sometimes that N=1 is pointing at something real. Something your competitors haven't noticed yet because they're too busy looking at aggregate data.

The Core Problem: Human Listening Doesn't Scale, and Subtlety Gets Lost

Here's where I was stuck as a product person before the last year or so:

I could recognize an edge case when I heard one. My pattern recognition was decent. But I couldn't listen to or review hundreds of them simultaneously. I had to rely on my team to funnel the "interesting" conversations to me. And by the time they did, context was lost. Emotion was flattened. The nuance—that subtle smile, that hint of something deeper—was gone.

So either you:

  • Have a small product with deep founder-to-customer relationships (hard to scale)

  • Rely on metrics and surveys (misses the signal entirely, especially subtlety)

  • Do quarterly research sprints (too infrequent to catch emerging patterns)

  • Hire a research team (expensive, and they still can't listen to everything)

Or you find a way to automate the listening without losing the authenticity and the subtlety.

That's the gap AI is starting to fill.

In Part 2, I'll walk through how AI is changing this dynamic—and why human sense-making remains irreplaceable (at least just yet).

#ProductStrategy #CustomerInsights #EdgeCases #FoundersJourney #ProductManagement #CustomerDiscovery #DataDriven