Why Most Redesigns Should Be Rewrites of the Content, Not the Layout
Clients ask for a new look. What they actually need is new words, new structure, and a reason to say less.
Every redesign brief starts the same way. The client sends over a competitor's site, some Pinterest boards, maybe a Figma file from an agency that never shipped anything. The ask is almost always visual: new colors, new type, a "modern" feel. Nobody asks for the thing that would actually fix the site, which is fewer, better, more honest words.
Bad layouts are rarely the root problem. They're a symptom of bad content trying to find a shape. A homepage with six competing headlines, three CTAs, and a hero section explaining what the company "believes in" isn't suffering from a design deficiency. It's suffering from nobody having decided what the page is actually for. You can restyle that page in any design language you want and it will still be confusing, just confusing with better kerning.
We've rebuilt sites where the client swore the problem was the navigation. Three rounds of nav redesigns later, the real issue surfaced: the site had four different products described in four different vocabularies, written by four different people over four years. No information architecture survives that. The fix wasn't a sitemap exercise. It was an editorial one — someone had to actually read every page, decide what mattered, and cut the rest.
This is why we push clients toward a content audit before a single wireframe gets drawn. Not a content inventory in the checklist sense, but an honest read: what is this page trying to make someone believe or do, and does every sentence on it serve that. Most pages fail this test badly. The layout can't save a page that doesn't know its own point.
There's a reason this gets skipped. Rewriting content is slower, less visual, and harder to put in a deck. Redesigning layout produces something you can screenshot and send to the board by Friday. But layout-first redesigns have a short half-life — six months later the site looks fresh and still doesn't convert, because the underlying confusion just got a new paint job. The sites that actually hold up are the ones where someone did the unglamorous work first: deciding what needs to be said, and only then deciding how it should look.