Rebranded or acquired a business? What to do with the old website.

The short answer

The old website does not die when the new one launches. It lingers in the index, in AI answers, in directories and in every link anyone ever gave it. The redirect map is how you hand all of that over to the new site instead of losing it or, worse, leaving two versions of the business competing with each other.

Why the old site keeps competing

Four things keep an old address alive long after you have moved on.

• Search engines remember URLs, not intentions. Until Google has recrawled an old address and followed a redirect, it can keep showing it.
• AI answers cite what they indexed. An answer assembled last month can still link to a page you retired this month.
• Listings outlive websites. In a sample we took on 27 September 2026 of the 26 companies GoodFirms lists as SEO companies in Guwahati, three of the listed domains no longer resolve at all and a fourth has become a domain-for-sale page. The directory still sends people to all four.
• Other people's links do not update themselves. Every article, profile and partner page that linked to you keeps pointing at the old address until someone changes it or a redirect catches it.

The same sample also had the right pattern in it. One company is listed under a .in domain that now permanently redirects to a .com, and the .com names itself as the canonical version. That is what a finished move looks like from the outside.

What Google says

Google's site move documentation is direct about the parts that matter.

Two more points from Google's redirect documentation change which redirect you use. A permanent redirect, 301 or 308, is used as a signal that the target should be the canonical URL. A temporary redirect, 302 or 307, is followed, but the indexing pipeline does not treat it as that signal. And where content has been deleted with no replacement, Google's guidance is to let the old URL return a 404 or 410 rather than disguise it.

Our own relaunch

We rebuilt this site on 2 August 2026, and the old URLs are still teaching us things.

• The relaunch removed the old service and blog sections. We wrote 27 redirect rules to hand them over, including both the slash and no-slash version of each address, because they are different URLs to a server.
• Six of the removed pages had only existed since 10 July, so they were live for 23 days.
• On 27 September, eight weeks after they were removed, two of those addresses were still listed in our Search Console generative AI report with three impressions each. The report cannot tell us whether those impressions happened while the pages were live or afterwards, and we are not claiming either.

What it does show is that a page can earn visibility in AI answers within weeks of existing, and still be on the books long after it is gone. The redirect is what decides where that visibility goes next.

What we got wrong

Checking our own map against Google's wording, two rules failed.

We had stopped offering two services, and we pointed both of their old pages at the homepage. That is the pattern Google names directly: many old URLs sent to one irrelevant destination. The homepage does not describe either service, so a visitor arriving from an old link or an old AI answer lands somewhere that does not answer what they came for, and Google may treat the redirect as though the page were simply gone.

On 30 September we removed those rules so the two addresses now return a plain 404. The page is gone and the response says so. The rest of our map passed: every other old URL points at a page that genuinely replaced it.

Build the redirect map

A spreadsheet with one row per old URL, and a decision in every row.

1. List every old URL. Pull them from the old sitemap, your Search Console pages report, your analytics landing pages, and any backlink tool you have. Pages that earned links or impressions matter most.
2. Give each one a decision. Same content on the new site: redirect to it. Content folded into a broader page: redirect there, if that page genuinely answers what the old one did. Content retired with no replacement: 404 or 410.
3. Use permanent redirects, 301 or 308, and one hop only. A chain of redirects is slower and gives every hop a chance to break.
4. Cover both versions of each address, with and without the trailing slash, and the www and non-www forms of the old domain.
5. Test from outside. Request every old URL and confirm the status code and final destination, rather than trusting the configuration file.
6. Keep the file for at least a year, and do not let a later redesign quietly drop it.

If you bought the business

An acquisition usually means two websites, two sets of listings and two names, and the job is to leave one of each.

• Decide which site survives before anyone touches the other. Keeping both live "for now" is how you end up with two businesses competing for the same searches.
• Map the absorbed site into the surviving one exactly as above, then submit a Change of Address in Search Console for the old domain.
• Say the former name in plain words on the surviving site, on the About page, so a person or a machine reading it can connect the two. "Formerly X, acquired in 2026" is enough. This is our recommendation, not a Google requirement.
• Merge or close duplicate Business Profiles rather than leaving two listings for one business at one address.
• Update every profile and directory that names the old business or links the old domain.

The first 90 days after

• Week one: request a sample of old URLs from outside and confirm every one lands where the map says.
• Weekly: watch the Search Console page indexing report for old URLs that are still indexed or unexpectedly 404.
• Monthly: ask the AI engines your buyers use about the old business name and see what they say and which address they cite.
• By day 90: every profile, listing and partner page you control points at the new address.

Limits

• Our relaunch kept the same domain. A move to a new domain adds the Change of Address step and usually takes longer to settle.
• The AI report shows impressions against old URLs but not when they happened, so it tells us the addresses were still on record, not that they were earning new visibility after removal.
• The Guwahati sample is 26 listings on one directory page, collected on one day. It shows that listings outlive websites. It does not measure how often.
• We describe the companies in that sample in aggregate only.