The Context
A client arrived convinced they needed a complete redesign. Their website was four years old, the design felt outdated, and competitors had more modern sites. Everything pointed to it being time for a change.
The Context
A client arrived convinced they needed a complete redesign. Their website was four years old, the design felt outdated, and competitors had more modern sites. Everything pointed to it being time for a change.
We agreed. The site looked dated. We proposed a ground-up redesign with updated technology.
It was a mistake.
What Seemed Obvious
The old website converted. Not spectacularly, but consistently. Users knew where everything was, the purchase flow worked, and SEO was established.
None of this came up in the initial conversation. We were all focused on visuals, on modernity, on what the competition was doing.
What Went Wrong
The new design was objectively better. Cleaner, faster, more current. Target users preferred it in tests.
But conversion dropped 23% in the first month. It rose slowly afterward, but never recovered to previous levels.
SEO suffered. Despite proper redirects and migrated content, we lost rankings that took months to regain.
Existing users complained. They had established habits that the new design disrupted. Some stopped buying.
Why It Happened
We confused old with bad. The previous site wasn’t bad; it simply wasn’t new. Those are different things.
We didn’t measure what mattered before making changes. We lacked a clear baseline for conversion by page, user flows, or real friction points.
We assumed better design automatically means better results. The correlation isn’t automatic.
What We Should Have Done
Measure exhaustively before touching anything. Understand what was working and why.
Question the premise. Did it really need a full redesign? Or targeted improvements at identified pain points?
Change incrementally. Test changes in limited sections before transforming everything.
Keep what worked. Don’t throw the baby out with the bathwater.
The Core Lesson
Change for its own sake isn’t progress. Before proposing a transformation, you must deeply understand what you’re transforming and why.
An ugly site that converts beats a beautiful one that doesn’t. Design serves business goals, not the other way around.
What We Do Differently Now
We start with data, not opinions. Before proposing solutions, we understand what the real problem is.
We question the need for big changes. Often the right answer is incremental improvement, not revolution.
We measure continuously. If we can’t measure the impact, we can’t know whether we’re improving or getting worse.
This project cost us a satisfied client and an expensive lesson. But we learned something more valuable: our intuition and the client’s aren’t enough.
Data doesn’t lie. Opinions do—including ours.