Plenty of products serve two populations who want opposite things. A first-time user wants fewer choices and a shorter path. An expert wants density, speed, and everything within reach. Give the beginner what they want and the expert leaves. Give the expert what they want and the beginner never gets started.
The obvious answer is to build both — a simple mode and a pro mode, or two apps. I’d push back on that, and the reason isn’t taste.
Two products is an operating decision
The moment you ship two versions, you’ve committed to two information architectures, two QA surfaces, two release paths and two backlogs, permanently. Every feature after that costs twice. The team quietly splits along the same line, and so does the roadmap.
None of that cost appears in the review where the decision gets made. It shows up eighteen months later as a velocity problem nobody can trace.
Depth on demand instead
The alternative is one structure where depth is available rather than default.
- The default path is the beginner’s. Shortest route, fewest decisions, nothing clever. Nothing they need is more than one step away.
- Expert tools are present but not preemptive. One gesture away, and they stay where they were left. An expert will happily learn a gesture; they won’t tolerate re-configuring on every visit.
- Both people learn the same product. The beginner isn’t being trained on a toy they’ll later have to unlearn.
The craft is all in sequencing — deciding what’s visible at rest and what earns its way onto the screen.
The test that settles it
If you’re arguing about this now, one question usually resolves it: what is the smallest thing a new user must understand before they can act?
Design for that, then make everything else discoverable. If the answer genuinely requires a second product, the problem isn’t the interface — it’s that you have two products already and haven’t admitted it.
What it bought
On a trading platform I led from a blank page, holding to one product meant a single backlog through a period when the app went past ten million downloads and its rating moved from 3.7 to 4.6.
I won’t claim the architecture caused all of that; plenty else changed. But I’ve watched teams take the split-product route, and I’ve watched them spend the next two years paying for it.