Most redesigns do not fail in the design work. They fail in the hand-off between “we need a new website” and the build starting: no agreed scope, so everything is apparently equally important; no record of what the old site was earning, so nobody notices what the new one lost; and no redirect map, so pages that took years to rank quietly point at a homepage or a 404. This guide assumes the decision is already made — if it is not, our 7 signs your website needs a redesign is the diagnosis step, and our redesign cost guide prices the work once the scope exists. This page is the execution layer between them: a scorecard that turns “redesign the website” into a prioritised, owned brief, the SEO baseline you capture before anything changes, and a redirect map template you can copy today.
How the scorecard works
Twenty items, five groups. Score each item 0 (not true today and not planned), 1 (partly true, or planned but unowned), or 2 (true, or planned with a named owner and a date). The groups are ordered by what they protect: money first, then the search traffic you already own, then content, then the technical build, then trust. The rules that make the score useful are three. A Must item at 0 blocks launch until it is fixed — these are the items whose absence is discovered after the damage. Should items go into the brief as scheduled work with an owner. Could items are written down and explicitly deferred, because an unwritten “could” is how redesign quotes quietly double. Score honestly: a 0 you admit now is a task; a 0 you discover at launch is an incident.
The scorecard
| # | Group | Item | Priority |
|---|---|---|---|
| 1 | Money pages | Your top 5 revenue pages (by enquiries or sales, not visits) are named in the brief | Must |
| 2 | Money pages | Every one of those pages keeps its URL, or has a redirect-map row | Must |
| 3 | Money pages | Forms, phone numbers and booking/checkout flows are tested on the staging site before launch | Must |
| 4 | Money pages | Analytics and conversion tracking are verified firing on the new site on launch day | Must |
| 5 | SEO preservation | The five-part baseline (next section) is exported and saved before build starts | Must |
| 6 | SEO preservation | A redirect-map row exists for every old URL — keep, redirect or retire, none undecided | Must |
| 7 | SEO preservation | No page that earns traffic or links is deleted without a named destination | Must |
| 8 | SEO preservation | Staging site is password-protected or noindexed, and the block is removed at launch | Must |
| 9 | Content | Each kept page has an owner for reviewing its copy in the new design | Should |
| 10 | Content | Pages that rank keep their topic, headings structure and search intent, even if rewritten | Should |
| 11 | Content | Thin or duplicate pages are merged deliberately (with redirects), not carried over by default | Should |
| 12 | Content | New photography/copy needs are listed now, not discovered during build | Should |
| 13 | Technical | Core Web Vitals targets are in the brief: LCP ≤ 2.5s, INP ≤ 200ms, CLS ≤ 0.1 on key templates | Should |
| 14 | Technical | The new site is checked on real phones (not just a resized browser) before launch | Should |
| 15 | Technical | Backups, security updates and who-does-maintenance are decided before launch day | Should |
| 16 | Technical | 404 page exists and offers the main routes back into the site | Could |
| 17 | Trust & UX | Contact details, opening hours and service areas match everywhere they appear | Must |
| 18 | Trust & UX | Reviews/testimonials that exist today are carried over, not lost with the old theme | Should |
| 19 | Trust & UX | Navigation is organised by what customers come to do, not by org chart | Should |
| 20 | Trust & UX | Accessibility basics: contrast, tap targets, alt text on content images | Could |
How to read the score. Maximum 40. 34–40: your brief is launch-ready; hand it over as written. 26–33: buildable, but close the Must zeros first and give every Should an owner before signing anything. 18–25: the redesign is still an opinion, not a plan — finish the baseline and the redirect map before discussing design. Under 18: do not brief a designer yet; you are about to pay for a prettier version of the same problems. Whatever the total, the Must column outranks it: a single Must at 0 — no baseline, a money page with no redirect row, tracking nobody has tested — is a launch blocker even in a 38-point project. When the scorecard is done, it is the brief: attach it to any quote request, and the quotes you get back will finally be pricing the same job. That scored scope is also what a website development team needs from you before honest timelines are possible.
Before anything changes: the five-part SEO baseline
The baseline is the evidence that lets you prove — to yourself, not just to Google — what the redesign did. It takes under an hour, it can only be captured while the old site exists, and it is the step every post-launch panic wishes had happened. Five exports, saved in one folder, dated:
- Search Console → Performance, longest range available, Pages tab, exported. Your traffic-earning pages and the queries they win. This is the list your redirect map must protect, in priority order.
- The same report, Queries tab. The words you actually rank for — the copy on kept pages has to keep answering these, whatever the new design looks like.
- GA4 → top landing pages, 12 months. Pages that start revenue journeys, which are not always the pages with the most traffic.
- A full crawl of the current site, exported. Every URL that exists today. Free crawler tiers cover a few hundred URLs, which is most small-business sites; this export is the raw material of the redirect map. Our technical SEO audit checklist goes deeper on reading a crawl if you want it.
- Search Console → Links, top linked pages. Pages other sites point at. A URL with backlinks that changes address without a redirect throws away the one asset money cannot quickly buy back.
Screenshot the homepage and your top three money pages while you are at it — not for SEO, but because “what did it look like before?” gets asked in every redesign, usually by the person who approved changing it.
The redirect map: template and rules
A redirect map is a promise, page by page, about where yesterday’s URLs live tomorrow. Build it from the crawl export, before the build starts, in a spreadsheet with these columns:
| Old URL | New URL | Action | Reason | Owner | Tested |
|---|---|---|---|---|---|
| /web-design.html | /website-design/ | 301 | Same service, cleaner URL | Dev | ✓ |
| /blog/2019-spring-offer | /offers/ | 301 | Dated post, offer page is the closest equivalent | Owner | ✓ |
| /old-location-page | — | Retire (404/410) | Location closed 2024, no traffic, no links | Owner | ✓ |
Five rules make the map work. Same URL first: if the new site can publish at the existing address, it should — a kept URL needs no redirect and loses nothing. Closest equivalent, never bulk-to-homepage: Google’s own guidance warns that redirecting many old URLs to one irrelevant page can be treated as a soft 404; a retired page goes to its nearest living relative or is honestly retired. One hop: old URL to final URL directly — chains of redirects slow everything and break silently. Permanent means permanent: use server-side 301 (or 308) redirects; a 302 tells Google the move is temporary and the old page should keep its place. Google’s site-move guidance asks that redirects be kept for at least 180 days; treat 12 months as the operating rule — redirects cost nothing to keep. Every URL gets a row: “keep” and “retire” are decisions too, and a URL with no row is a URL nobody decided about. If your redesign also changes platform or domain, the stakes rise again — our website migration guide covers that case end to end, and note that Google’s Change of Address tool is for domain moves only, not for tidier URLs on the same domain.
Launch week, and the 30 days after
Launch day, in this order: confirm the staging block (password/noindex) is removed; submit the new sitemap in Search Console; spot-test the redirect map’s top 20 rows — every money page and top traffic page — by actually visiting the old URLs; confirm analytics and conversion tracking are firing (scorecard item 4); click every form and the checkout once, for real. Days 2–30: once a week, compare Search Console clicks and impressions against your baseline export, and check the Pages report for new “not found” or “redirect error” entries — those are map rows you missed, and they are fixable in minutes while the baseline still tells you what the page used to earn. Expect movement: Google’s own position is that rankings shift after a redesign because the pages themselves have changed, and some movement cannot be prevented by any checklist. The baseline is what separates that normal settling from real damage — a page that held position for a year and fell off a cliff the week of launch, with a broken or missing redirect, is damage; a two-position shuffle across the board is settling. Keep the redirects, keep watching for the full 30 days, and do not redesign the redesign in week two.
Platform branches
The scorecard is platform-neutral; three steps change shape by platform. WordPress: redirects live in your host’s configuration, a redirection plugin, or .htaccess — pick one home and keep the map as the source of truth, and test after any plugin update in launch month. Shopify: URL structure is fixed in places (/products/, /collections/), so “same URL first” often means accepting Shopify’s structure and mapping to it; Shopify’s built-in redirect manager handles the rows, but only for paths Shopify no longer serves. Website builders (Wix, Squarespace and similar): redirect support varies and is sometimes page-count-limited — check your plan’s redirect capacity against your map’s row count before the build, not after; if the map is bigger than the tool, that is a Must-level finding about the platform, not a detail.
FAQ
When should the redirect map be built — before or after the new site is designed?
Before the build starts. The map is built from the old site’s crawl and baseline, and it constrains the new site’s URL structure; built afterwards, it becomes a list of apologies. If URLs are still changing during the build, keep the map open as a living document — but the first complete draft exists before development.
How long should redirects stay in place?
Google’s guidance asks for at least 180 days. Our operating rule is 12 months minimum, and indefinitely for any URL with backlinks — a redirect costs nothing to keep and protects traffic and links for as long as it stands.
Will following this process guarantee my rankings do not drop?
No, and any provider who promises that is selling something. Google’s stated position is that some movement after a redesign is normal and cannot be fully prevented. What the scorecard, baseline and map do is remove the unforced losses — lost pages, missing redirects, broken tracking — and give you the evidence to tell normal settling from real damage within days instead of months.
The short version
Score the scope before you brief it: twenty items, Must items at zero block the launch. Capture the five-part baseline while the old site still exists — it is the only evidence you will ever have of what the redesign changed. Build the redirect map from the crawl, one row per URL, same address wherever possible, closest equivalent where not, 301 and one hop. Then launch carefully and watch Search Console against the baseline for 30 days. If you would rather run the redesign with a team that works this way, our website development service starts every rebuild with exactly this scorecard — get a quote and bring your scores.
A note on what this article does and does not claim
The scorecard, its scoring bands, the baseline gate and the watch sequences are this article’s own working method, written for small sites of roughly 10–100 pages; they are a framework, not measured results, and no redesign outcomes were studied for it. The redirect mechanics stated — server-side 301/308 as permanent, 302 as temporary, Google’s soft-404 warning about bulk redirects to one page, the at-least-180-days guidance, and the Change of Address tool being for domain moves — are Google’s published site-move guidance, read for this article via two independent guides quoting that documentation rather than re-fetched from Google’s pages in this pass; the 12-month redirect rule is our recommendation, not Google’s figure. The Core Web Vitals thresholds are Google’s published “good” thresholds. Google’s position that some ranking movement after a redesign is normal and cannot be fully prevented is stated in the article because it is true, and it applies to this process too: the method removes unforced losses, it does not and cannot promise rankings will not move. This article deliberately contains no redesign prices — that intent belongs to the cost guide linked above.
— AI Dev Expert. Researched and drafted with AI assistance; reviewed against those sources.
