SEO before a website rebuild: what to preserve to protect traffic

Why a rebuild can reduce traffic

Search engines evaluate individual URLs and their content, not only a company name. An old URL may be associated with external links, indexing history, user queries and conversions. When implementation changes directory structure, removes pages or replaces detailed content with a short description, both the crawler and user reach something different.

This is not an argument against change. A planned redesign can improve speed, mobile usability and clarity. Treat it as an information migration: decide what stays, what gets a replacement, and how the result will be verified.

Start with inventory, not design

Before wireframes, create one table. For every public URL, answer: “what happens to this address after launch?” Gather URLs from Search Console, analytics, sitemaps, navigation, campaigns, social profiles and external links.

Do not mix three different decisions

1. Same URL, better page

This is the safest option when the page topic and intent remain the same. Keep the URL, improve hierarchy, mobile usability and content, then test canonical, indexability, forms and links.

2. New URL for the same topic

Use a permanent server-side 301 redirect from the old URL directly to the closest relevant new page. Avoid redirect chains. Do not automatically send every retired page to the homepage: for users it is often a dead end, and for search engines it is a weak relevance signal.

3. No replacement page

Removal can be correct when a service or material no longer exists. Record the decision and check whether navigation, articles, campaigns or business profiles still link to it.

What to preserve beyond URLs

For key pages, record the existing title, H1, meta description, important subheadings, service information, contact details and trust signals. The aim is not mechanical copy-paste; it is to preserve the topic, evidence and information that made the page useful.

Also check canonical URLs, index directives, HTTP statuses, sitemap and internal links. A page available on multiple URLs or accidentally marked noindex can be hard to index even when its content is good.

A practical seven-day plan before launch

1. Day 1: export Search Console and analytics data; mark pages with traffic, conversions and links.
2. Day 2: make the old URL → new URL map and assign an owner to every unclear decision.
3. Day 3: compare key content and confirm that offer, contact and legal information remain available.
4. Day 4: configure and test redirects in staging; one old URL should lead to one final URL.
5. Day 5: test navigation, CTAs, forms, links and horizontal overflow on a phone.
6. Day 6: prepare an acceptance list: 200 for destination pages; 301 for migrations; canonical, title, H1 and index,follow.
7. Day 7: repeat the list after launch and update the sitemap with canonical URLs only.

Acceptance check after launch

Do not end with “the website works”. Open the URL map and check every address. A retained URL should return 200 with correct content. A changed URL should return one 301 redirect and a 200 destination. On the destination, confirm one relevant H1, self-referencing canonical, no noindex, working internal links and a visible CTA.

What Google recommends checking

Google Search Central’s guidance on URL migrations recommends preparing a URL map, using permanent redirects and monitoring traffic after change. On the same domain, correct destination URLs, redirects and indexability remain central.

Source: Google Search Central — Site Moves and Migrations.