Rebranding: Why AI Keeps Citing Your Old Company Name

🇫🇷 Version française

You changed your company name months ago, redid the logo, told your customers. Yet ask ChatGPT or Perplexity about your business and the old name still comes up — sometimes alone, sometimes mixed with the new one. This isn't a glitch. It's the logical result of how an AI actually builds its knowledge of your brand.

A rebrand is an internal project: new logo, new guidelines, new website, an announcement to existing customers. But from an AI's perspective, none of that instantly erases the old name. That name still lives on in thousands of pages you don't control: press articles, professional directories, customer reviews, old social media posts, mentions on partner websites. As long as those traces exist and aren't explicitly linked to your new name, an AI keeps treating them as valid sources.

Why does AI take so long to "forget" your old name?

A generative AI doesn't check in on your business live for every question: it draws on a set of sources collected at different points in time, some very recent, some older. If the old name appears across a large share of those sources — and the new one in very few — the statistical odds mechanically tilt toward the old name, even if it's no longer accurate.

So it's not that the AI "prefers" your old name. It's that, from where it stands, nothing yet proves at scale that the new name is the official, current continuation of the same business. Without that explicit link, repeated across several trustworthy sources, the two names can even be read as two separate entities.

Which sources keep pulling the old name back into answers?

Some sources carry more weight than others in keeping the old name alive:

• Professional directories and local listings — many are never updated after a name change, and remain sources treated as reliable for practical information.
• Press articles and past interviews — impossible to correct retroactively, they keep associating your business with the old name.
• Existing customer reviews — a customer who left a review under the old name a year ago isn't going to update it on their own.
• Your own old pages still indexed — if old URLs or versions of your site remain accessible without a clear redirect to the new name, they feed the confusion instead of resolving it.

How to speed up the switch without rewriting everything

The goal isn't to erase the old name — that's usually impossible, and sometimes counterproductive if existing customers still search for it. The goal is to multiply the reliable, recent sources that explicitly connect the old name and the new one, so an AI no longer has to guess.

In practice, that means updating the sources an AI checks first to identify you: your website, your Google Business Profile listing, your professional and social profiles, and any directories or third-party databases that still mention you. A dedicated page that clearly states "formerly known as [old name], we are now [new name]" — published on your own site and echoed elsewhere — gives the AI the explicit bridge it needs to stop hesitating.

Should you worry if the old name still comes up?

No need to panic, but don't ignore it either. The real risk isn't that an AI mentions your brand's history — it's that it does so incorrectly, or worse, cites the old name as a still-separate entity from you, splitting your authority between two identities instead of one. That's the case worth actively correcting, not a well-clarified mention of your history.