The expert user death spiral
Some products end up usable only by experts not because anyone decided that, but because of a slow, self-justifying slide. Each step feels reasonable on its own. Strung together, they trap a product in a shrinking market it mistakes for loyalty. I’ve come to think of it as the expert user death spiral, and it runs in four stages.
Stage I: “Some expert knowledge required”
In the beginning, complexity is honest. The product solves a genuinely hard problem, and the first people through the door are knowledgeable. They understand the domain, they’ve felt the pain you’re solving, and they’ll tolerate rough edges because the value is worth it. A steep learning curve isn’t a flaw here; for early adopters it can even read as a sign of depth.
This stage is fine. The trouble is that it sets a precedent. The team learns that users will climb the curve, and that lesson quietly hardens into an assumption.
Stage II: “All our users are expert users”
Now the belief takes hold. Because the people using the product are expert, the team concludes that the people who use the product must be expert, and always will be. Onboarding gets deprioritised. Awkward flows survive because “our users know the workaround.” Usability concerns are waved away with “anyone who’d use this already understands it.”
This is survivorship bias wearing a lab coat. You’re only looking at the users who made it past the curve. The ones who bounced off (and they might represent a far larger market) left no trace in your analytics and no seat in your meetings. Their absence looks like proof that non-experts don’t want the product, when it’s really proof that the product didn’t let them in.
Stage III: The product selects for experts
Belief becomes mechanism. A product that assumes expertise turns non-experts away at the door, so the only people who stick around are, inevitably, experts. The user base self-selects to match the assumption, and every cohort review confirms it: “see, our users are all power users.” The evidence and the belief now reinforce each other on a loop:
Technical product → technical users → usability deprioritised → more complexity → fewer non-technical users → “all our users are experts.”
This is the same loop I described in Design Debt vs. Tech Debt, but here it’s worth naming what it does to the team, not just the product. It launders an accident of who-showed-up into a deliberate-sounding strategy. Nobody chose to build an experts-only product; the spiral chose it for you, and then handed you the data to feel good about it.
Stage IV: The trap
Eventually growth flattens, because you’ve saturated the small pool of people willing to climb the curve. Someone, often someone new, often from outside design, asks the obvious question: why don’t we open this up to a wider, less technical market?
And you discover you can’t, at least not cheaply. The complexity is woven through the information architecture, the data model, the workflows, and the muscle memory of every existing user, far beyond any one screen you could redesign over a sprint. Simplifying now means either an expensive rebuild or a long, disruptive migration for the very experts who kept you alive, the kind of careful, incremental surgery I wrote about in how to update a complex legacy application. The growth you wanted was available years ago and is now locked behind a wall you built one reasonable decision at a time.
Breaking the spiral
The spiral is powered by a single unexamined belief, so that’s where you break it.
- Distrust “our users are all experts.” Treat it as a hypothesis to test, not a fact to plan around. Where did that belief come from, and who isn’t in the room because of it?
- Go looking for the people who left. Talk to users who abandoned onboarding, or who evaluated the product and chose something else. They are the market you can’t currently see.
- Benchmark usability against non-experts. A System Usability Scale score from your existing power users tells you almost nothing about reach. Test with people who aren’t already fluent.
- Decide expertise on purpose. There’s nothing wrong with building a deliberately expert tool; some of the best products are exactly that. The danger is backing into it by accident and only noticing once it’s expensive to leave.
Comfort with a loyal expert base feels like product–market fit. Sometimes it is. But if it’s the death spiral instead, the bill doesn’t arrive until the day you try to grow, and by then it’s a rebuild rather than a redesign.