I keep landing inside products that work. The metrics aren’t screaming, support isn’t on fire, the navigation mostly holds. And still, the experience feels heavy, like walking through a house that’s clean but crowded. That feeling is why I’m writing this—to remind myself (and anyone who needs it) that a redesign is rarely the medicine. The cure is a reckoning.
What “off” really feels like
“Off” is when the interface says look at me while the user says let me go. The screens are busy explaining themselves. Microcopy props up choices that should be self-evident. A tooltip sits where a better control ought to be. Nothing is truly broken, yet everything requires just a little too much decoding. I can use it, but I have to think about it—and thinking is the tax.
Redesign or reckoning
I don’t reach for a redesign when the product feels off. I reach for sharper questions. A redesign comforts the eye; a reckoning clarifies intent. One is paint; the other is carpentry. Paint is wonderful, but if the doorframe is crooked you’ll still feel it every time you walk through.
The questions I ask first
- If I had to explain this navigation at the kitchen table, could I do it in a sentence per item?
- Which elements have not earned their place in the last twelve months—by usage, by outcomes, or by common sense?
- Where does the language betray the business we used to be? (Old plan names, MVP-era verbs, promises we no longer make.)
- Which actions describe tasks rather than outcomes? ("Create report" versus "See your cost breakdown".)
- Where are the dead ends—the buttons that lead to waiting rooms, the paths that don’t end in a result?
I don’t need perfect data to begin; I need honest eyes and a willingness to remove.
The gentle brutality of subtraction
When I’m brave, I delete first. I demote features with under 2% usage. I archive settings that mirror internal systems but not user jobs. I replace three steps with one even if it means upsetting a familiar ritual. I’d rather be naïve than cynical: if a thing adds friction without adding confidence, it can go. If we miss it, we can return it with better intent.
Language as navigation
Language is cheap and powerful. I pull verbs forward. I replace names that sound like codenames with words people actually use. I move proof to where the decision happens—a short stat beside the button instead of a testimonial page no one visits. I label states openly: Generating… → View report. It’s mundane work. It also changes behaviour.
The kitchen-table test
I imagine a friend sitting across from me: no slides, no Figma. Can I say the journey in one breath? Land → Learn → Try → Value → Convert. If I can’t, the product probably can’t either. Wherever I stumble, I cut or rename. Complexity that hides in language will ambush the user in clicks.
A one-week promise to myself
Day 1: pull baselines—activation, time to first value, the top three support themes.
Day 2: inventory the surface—nav, key screens, modals, settings; note owners and last updated.
Day 3: remove or demote anything unused, duplicated, or orphaned.
Day 4: rewrite labels and CTAs to reflect outcomes.
Day 5: place proof where decisions are made; fix empty states so they teach one next step.
Day 6: watch sessions or call five users; fix what confused them once.
Day 7: ship. Measure for fourteen days. Don’t argue with the numbers.
Why this is hard
Audits don’t flatter anyone. They expose favourite ideas that never quite landed and patterns we copied because we were tired. They ask for ownership—of screens, of states, of words. Redesigns, by contrast, feel like renewal. New colours, new type, a sense of progress. But repackaging confusion is still confusion.
What relief feels like
Relief is the moment a screen stops introducing itself and starts doing its job. The primary action is obvious without shouting. The empty state teaches. The button tells the truth about what will happen next. You can feel the product saying this way and you don’t have to ask why.
My bottom line
If it feels off, I start small and sharp. Remove what doesn’t belong. Clarify what does. Align words with outcomes. Put proof beside decisions. Then, maybe, I’ll talk about a redesign—once the structure is honest. Looking better by working better is not romantic. It’s the adult choice.
That’s the note I needed to write. If you needed it too, try a week of reckoning. The product—and the team—will breathe again.
Top comments (0)