Against Redesigns (From Someone Who Does Them)

I redesign websites for a living. So believe me when I say: most redesigns shouldn’t happen.

Not because the current site is fine. Usually it isn’t. But because a redesign — a full tear-down-and-rebuild — is almost never the right response to what’s actually wrong.

The Redesign Impulse

A business owner looks at their website and feels embarrassed. Maybe a competitor launched something slick. Maybe a friend made a comment. Maybe they just haven’t touched it in three years and the aesthetic has drifted from “current” to “dated.”

The instinct is: burn it down, start fresh.

This instinct feels decisive. It feels like progress. But it’s usually a way of avoiding the harder question: What specifically isn’t working?

“The whole thing” is never the real answer. Dig deeper and you’ll find the actual problems are specific:

The homepage doesn’t explain what we do
Mobile users can’t find the phone number
The booking form is buried three clicks deep
Page speed is terrible because of that background video
We’ve added four new services and they’re not on the site

These are refinement problems, not redesign problems. You can fix each one in a day. A redesign takes months.

What Redesigns Actually Cost

The sticker price is the smallest part.

Lost SEO equity. Every URL change, every restructured page, every renamed heading risks dropping search rankings. Google has indexed your current site. A new site is, from Google’s perspective, a new site. Even with perfect redirects, there’s typically a 3-6 month recovery period.

Content migration pain. Your current site has years of accumulated content — blog posts, product descriptions, testimonials, FAQs. Migrating this to a new structure is tedious, error-prone, and almost always underestimated.

Staff retraining. If there’s a CMS, everyone who edits the site needs to learn the new system. This is invisible cost — it shows up as slower content updates for months after launch.

The “while we’re at it” trap. Once a redesign is approved, every stakeholder adds their wish list. “While we’re rebuilding, can we also…” The scope doubles. The timeline triples. The budget becomes a polite fiction.

When Redesigns Are Right

Sometimes the foundation is genuinely broken. Signs that a real redesign is warranted:

The tech stack is abandoned. If your site runs on a platform that’s no longer maintained, security patches stop coming and you’re on borrowed time.

The business has fundamentally changed. You used to sell products, now you sell services. You used to serve local customers, now you serve national ones. The site’s entire information architecture is wrong, not just the styling.

Performance is structurally unfixable. Some sites are so bloated with plugins, trackers, and legacy code that optimising within the existing system costs more than rebuilding.

The brand has changed. New name, new positioning, new visual identity. The site needs to reflect a genuinely different business, not a tweaked version of the old one.

If none of these apply, you probably need a refinement, not a redesign.

Instead of rebuilding, iterate:

Month 1: Fix the three biggest usability problems. Usually these are obvious — broken mobile layouts, missing calls to action, slow load times.

Month 2: Rewrite the homepage copy. This single change often has more impact than any visual redesign. Say what you do, who you do it for, and why someone should care.

Month 3: Update the visual design incrementally. New typography, refined colour palette, better photography. Applied to the existing structure, not a new one.

Month 4: Add what’s missing. The new service page. The portfolio section. The FAQ that would reduce support emails by half.

Four months of focused refinement, each change measurable, each building on the last. No big-bang launch. No holding your breath hoping nothing breaks.

The Honest Conversation

When a client asks me for a redesign, I ask them to show me five specific things they’d change. Not vague feelings — specific elements on specific pages.

Usually they can name three. Sometimes five. Rarely more.

If the list is short and specific, I suggest refinement. It’s faster, cheaper, lower risk, and the client ends up happier because we solved their actual problems instead of creating a whole new set.

If the list is long and structural — “the whole navigation is wrong,” “none of the content reflects what we do anymore,” “we can’t update anything without calling the developer” — then yes, we rebuild. But we rebuild with a clear brief, not a vague desire for “something better.”

The Irony

The best redesigns I’ve done are the ones where the client couldn’t tell I’d redesigned anything.

The layout shifted subtly. The typography improved. The load time dropped. The contact form moved above the fold. The booking flow lost two unnecessary steps.

The client’s customers didn’t notice the change. They just found the site easier to use. That’s the point.

A redesign should be invisible to the end user. If visitors notice your new website, you’ve prioritised your ego over their experience.

The best redesign is the one your customers never notice.