Blog Overview
A URL inventory, a 301 map, and eight checks for the first 72 hours. The checklist we use before and after moving a domain, and the four causes of traffic loss.
A website redesign starts with enthusiasm and often ends with a traffic drop. Not because the design is bad, but because every old URL that used to bring a visitor is now a 404. The drop does not appear on launch day. It appears two weeks later, when search engines re-crawl and the team is busy with launch instead of migration.
Migration is not a technical step at the end of a project. It is a content and information-architecture decision made before the first template. This article gives you the checklist we use: what to inventory before design, how to build the redirect map, and what to check in the first seventy-two hours.
Where sites lose their traffic
Four causes cover most drops. First, URLs without redirects: old service and article pages disappear. Second, a URL structure change without a map: the path still works, but the address changed. Third, deleting content that was bringing traffic because someone decided it was “old.” Fourth, slowing the site with add-ons and effects that were never tested on a phone.
Note that three of the four are content decisions, not technical ones. The engineer executes the map, but does not know which page deserves to survive. That knowledge sits with whoever wrote the content and whoever read the search reports, and they are usually not in the launch room.
The inventory before design
Start with one file containing every URL on the domain, with sessions over the last twelve months, the page that links to it, and the commercial purpose. Pull the list from the sitemap, from search console, and from server logs. Do not rely on the sitemap alone, because it usually omits old pages that are still visited.
Classify every URL into one of five states. The page that carries demand and needs improvement. The page that stays as it is because it works. The page that merges into a stronger one. The page that is deleted and redirected. The page that does not exist yet and should be created. This classification is what defines the redirect map later.
The redirect map
One rule: do not redirect everything to the homepage. A homepage redirect tells the search engine the old page is gone, and tells the visitor you do not care what they asked for. Send each URL to the closest page that carries the same intent, and if none exists, create it.
| Old URL type | Correct destination | What you lose by getting it wrong |
|---|---|---|
| Service page | The new service page with the same intent | Direct commercial demand |
| Article or guide | The closest new article on the topic | Informational traffic and external links |
| City or branch page | The general service page or contact page | Local search matching |
| Removed product page | The replacement or parent category | Near-purchase conversions |
| Legacy or intermediate page | The new parent page | Internal link equity |
From reading to shipping
Want this built, not just explained?
Bring us the problem behind the article. We will tell you what it takes, what it costs, and whether we are the right team.
Test the map before launch in a staging environment. Every redirect should return 301, not 302, and should land on a 200 page in one hop. A three-step chain loses part of the signal and slows the visitor. A redirect to a 404 is worse than no redirect, because it tells the search engine you knew about the page and abandoned it.
The first seventy-two hours
After launch, crawl behaviour changes quickly. These are eight checks ranked by impact, and each takes minutes rather than days.
Open the new sitemap. Confirm every URL returns 200 and the count matches what you expect.
Test ten high-traffic old URLs. Each should land on a relevant page in one hop.
Submit the new sitemap. In search console, and request indexing for the core pages manually.
Watch 404s in the logs. Any URL appearing more than ten times needs a redirect the same day.
Compare titles and descriptions. Confirm key pages did not lose their title or inherit a default one.
Check mobile speed. LCP, INP, and CLS on a slow connection, not a fast desk line.
Rebuild the internal link graph. Internal links should point to new URLs, not old ones.
Keep the logs. Compare the first week to the previous month, not the previous day, to separate noise from a real drop.
Content that should actually be deleted
Not every old page deserves to survive. Duplicate pages, generated content with no value, and pages that serve no clear intent all consume crawl budget and weaken quality. Delete them, but redirect their URLs to the closest alternative. Deletion without a redirect is the one mistake that cannot be repaired months later.
If you are building a new site from scratch, read the published cost guide on corporate website cost in Saudi Arabia, because it explains where migration sits inside the scope. To see how we run this work, see web platform development.
Frequently asked questions
Do we keep old URLs forever? No. Redirects stay as long as they bring traffic or external links. Review them yearly and remove unused ones.
Is an .htaccess file enough? It is the execution tool, but the map is the decision. Write the map in a spreadsheet first, then implement it.
Do we lose rankings if we only change the design? Usually not, if URLs and content stay. The risk comes from changing titles and structure.
How long does recovery take? It depends on site size. What matters is weekly monitoring and fixing drops on the day they appear.
Do we need an agency to migrate? No, but you need someone who owns the map and knows the content. That role is not purely technical.
Sources
- Google Search Central: site moves with URL changes — 301 handling and sitemaps.
- web.dev: Web Vitals — the performance thresholds we check after launch.
- WCAG 2.2 — the accessibility criteria a migration must not lose.
- Web platform development at Rasad — how we build and migrate.
