How to Plan a Website Migration Without Losing Traffic

Migrations lose traffic for one reason far more than any other: URLs changed and nobody mapped them. Everything else — platform choice, new design, improved speed — is secondary to that single piece of work.

Inventory before anything else

Crawl the existing site and export every URL. Then add URLs from Search Console, analytics and your sitemap, because a crawl finds only what is linked and some valuable pages are not.

For each URL, record its traffic, its rankings and whether it has inbound links. This inventory is what tells you which pages matter, and it is the document the entire migration plans around.

Do this before design decisions are made, because knowing that four pages produce sixty percent of organic enquiries changes what the new site must preserve.

Decide what happens to every URL

Each existing URL gets one of four outcomes: keep the same address, redirect to a specific new one, merge into another page, or retire. Retiring should be rare and deliberate.

Keeping a URL is always safer than redirecting it. Change addresses only where the existing structure is genuinely wrong, not because the new platform's default pattern differs.

Redirect to the closest relevant page, never to the homepage. Homepage redirects are treated much like a 404 and are the most common shortcut taken under deadline pressure.

Carry the content and the metadata deliberately

Titles, descriptions, headings and structured data should be migrated rather than regenerated. Pages ranking well are doing so partly because of how they are written, and a well-meant rewrite during migration is a common cause of losses that get blamed on the platform.

If content needs improving, do that after the migration has settled, so cause and effect stay separable.

Rehearse on staging

Run the full migration into a staging environment first. Check record counts, sample pages across every template, confirm redirects resolve correctly, and verify that content and metadata arrived intact.

Make sure staging is blocked from indexing while you do this. A staging site accidentally indexed during a migration creates duplicate content at exactly the wrong moment.

This rehearsal is where problems are meant to be discovered. Skipping it moves the discovery into production.

Launch deliberately and watch

Choose a low-traffic window with people available to monitor, and not immediately before a weekend. Have a rollback plan and confirm you can execute it.

Immediately after launch: submit the new sitemap, check index coverage daily for the first week, watch for crawl errors, and test the redirect map against the original URL list rather than assuming it worked.

Expect a short dip while search engines process the change. A dip that does not begin recovering within a few weeks means something is wrong and needs investigating, not waiting out.

Two things worth separating

Migrating and redesigning simultaneously saves money and makes diagnosis harder, because when traffic moves you cannot tell which change caused it. Where the site is commercially important, separating them is worth the extra cost.

Similarly, avoid changing the domain and the platform at the same time unless there is no alternative. Each is a manageable risk; together they compound.