Redesign is surgery, not fashion
Your existing site earned things: rankings, backlinks, indexed URLs, a user base that knows where things are. A redesign done carelessly sets all of it on fire to get a fresher look. The discipline is simple to say and rare to see: change what the business needs changed, keep what the site earned, and map every difference deliberately.
Step 1 — Audit before you redesign
Before a single mockup: what does the analytics say? Which pages drive conversions, which drive nothing, which URLs have backlinks and rankings worth keeping? The audit decides what "redesign" should mean — and a third of the time it says the site needs a refresh and specific fixes, not a rebuild. That finding is worth real money; it is the cheapest line item in the project.
Step 2 — Choose the right depth
| Depth | What changes | Typical band |
|---|---|---|
| Refresh | Design, copy, speed — same structure and URLs | $2k–8k |
| Structural redesign | Information architecture, templates, some URL changes — full redirect map required | $5k–20k |
| Replatform | New CMS/platform underneath — a migration wearing a redesign's clothes | $8k–40k |
Most businesses that say "redesign" mean the first row, are quoted for the second, and fear the third. Naming the depth is half the decision.
Step 3 — Protect the assets
- URLs: every URL that changes gets a 301 redirect to its mapped successor — decided before build, tested before launch. This one discipline is the difference between keeping your rankings and a 404 apocalypse.
- Content: the pages that convert move forward, improved — not deleted by aesthetic. Content audits come before design, or design becomes content destruction.
- Tracking: analytics continuity across the cutover, or you lose the ability to measure what the redesign did — the only way to know if it worked.
- Search Console: verified before, watched daily after. Indexing errors caught on day two are trivia; caught on day sixty they are history.
Step 4 — Do not launch on a Friday
Launch rehearsed: staging signed off against the launch checklist, rollback plan written, a weekday morning cutover with both parties watching. The sites that survive launches uneventfully are the ones that rehearsed them.
The uncomfortable question the guide exists for
Would a visitor who cannot see your design — a search engine — notice this project happened? If the honest answer is "no redesign was needed, the traffic problem was speed and content", then that is the project. If the answer is "yes — structure, speed and conversion paths all change", run it properly: audited, mapped, rehearsed. Briefs here start with that audit — including the one that ends with "you need less than you asked for".