How Do AI Assistants Handle a Product Rename?

Will AI assistants pick up your new product name right away?

Usually not. AI assistants learn about products from training data, which can be months old, and from pages they retrieve live, which may still use the old name. After a rename, assistants often mix the two, describe the old product, or treat them as separate tools. You need to help them connect the names on purpose.

Product renames are common in B2B software. A company repositions, merges two products, expands beyond its original category, or hits a trademark issue. Most rename plans cover the website, the app, sales decks, and email. Very few cover how ChatGPT, Claude, Gemini, Perplexity, or Google's AI Overviews will describe the product afterward.

That gap matters because buyers now ask AI assistants about vendors before they ever visit a website. If an assistant says "I am not aware of that product" or describes the old version, you lose credibility at the very first touch.

Why do AI assistants cling to old product names?

Because the old name is everywhere the assistant learned from. Years of reviews, articles, forum posts, docs, and directory listings use it. A model trained on that material links your company to the old name strongly. A few weeks of pages with the new name are a small signal next to that history, especially in training data.

Two different mechanisms are at work. A model's built-in knowledge reflects its training data up to a cutoff, which I explained in more detail in what a knowledge cutoff means for outdated AI answers. Retrieval, where an assistant searches the web while answering, can pick up new pages faster, but only if those pages are findable and clear.

So your job has two parts. Make the connection between old and new names explicit everywhere you control, so retrieval finds it. Then work on the wider web, so future training data reflects the change too.

What should your own website say after a rename?

State the rename plainly in text. Your homepage, about page, and main product page should say something like "Product New, formerly Product Old" for a period after the change. Publish a short announcement page that explains what changed, why, and what stayed the same. Plain sentences are what assistants retrieve and quote.

Many companies hide the old name immediately because they want a clean break. That is understandable for branding, but it removes the bridge that both people and machines need. A buyer who searches the old name should land on a page that explains the new one.

Keep the "formerly" wording in the about page and the announcement page long term, even after you remove it from the homepage. Those pages become the reference that connects the two names for years.

How should you handle URLs and redirects?

If old product pages or a domain carried the old name, redirect each old URL to its closest new equivalent with permanent 301 redirects. Do not send every old URL to the homepage. One-to-one redirects keep the meaning of each page intact, which helps search engines and AI retrieval follow the change accurately.

Map redirects before you launch the rename, not after. A rename often comes with a site update, and broken or chained redirects are easy to create in the rush. My guide on building a redirect map before a site migration walks through the process I use on Webflow sites.

Keep the old domain registered and redirecting for the long term. Old links in articles, reviews, and docs will keep sending traffic, and each one is a chance to show the new name.

Does schema markup help with a rename?

It can. Schema.org defines an alternateName property as "an alias for the item", and it can be used on any type. Adding the old name as an alternateName on your Organization or product markup gives machines a structured, explicit link between the two names, alongside the plain-text "formerly" sentence on the page.

Schema is a support signal, not a fix on its own. It works best when the visible text says the same thing. If your markup claims an alias that appears nowhere on the page, you are relying on a signal that systems may treat with less weight than plain content.

If you have not added Organization schema yet, the rename is a good moment. I covered the basics in adding Organization schema to a Webflow site.

What should you update outside your website?

Update every profile you control: LinkedIn, review sites, marketplace listings, integration directories, app stores, and partner pages. Ask partners and customers who wrote about you to update their pages where possible. These third-party sources can strongly shape how AI assistants describe a company, so leaving them on the old name keeps the confusion alive.

Make a list before you launch. Search for the old name, note every page that mentions it, and sort them into ones you control, ones you can ask about, and ones you cannot change. Work through the first two groups in the first month.

Your own docs and help center count too. Old screenshots and old names in documentation are a common source of mixed answers, because assistants often retrieve docs when users ask how a product works.

How do you check what AI assistants say now?

Ask them directly, on a schedule. Run the same questions through ChatGPT, Claude, Gemini, and Perplexity every few weeks: what is your product, what was it formerly called, and how does it compare to competitors. Record the answers. Look for old names, wrong descriptions, or confusion between the two.

Keep the questions identical each time, so you can see change rather than noise. Note whether each assistant cites a source. If an assistant cites an outdated page, that page is your next fix.

Expect the change to be gradual. Retrieval-based answers may update within weeks once your pages are clear. Built-in knowledge may take until models are retrained, which you cannot control. Focus on what you can control and keep measuring.

What mistakes make rename confusion worse?

The worst mistakes are deleting old pages without redirects, removing every mention of the old name, using different new names in different places, and announcing the rename only in a press release. Each one breaks the bridge between old and new. Consistency across your site, profiles, and docs matters more than any single tactic.

Inconsistent naming is surprisingly common. The website says one thing, the app says another, and the sales deck uses a shortened version. Assistants pull from all of these and may treat them as three products.

Pick one exact new name, including spacing and capitalization, and use it everywhere. Put it in a short naming guide for your team so new pages stay consistent.

How long should you keep mentioning the old name?

Keep a clear "formerly" mention on key pages for at least a year, and keep it on your about page and rename announcement indefinitely. Buyers, analysts, and AI assistants will encounter the old name in older content for years. A permanent, clear reference costs nothing and keeps answering the question.

You can reduce how prominent it is over time. Lead with the new name everywhere, and move the old name into a short line rather than a headline. But do not remove it entirely just because the rename feels like old news inside your company.

Inside the company, a rename feels finished the day it launches. Outside, it is only beginning. Plan for that difference.

What should you do next?

If a rename is coming, add an AI visibility step to the plan: "formerly" wording on key pages, a rename announcement page, one-to-one redirects, alternateName in your schema, and updates to every profile you control. If the rename already happened, ask the major assistants about your product today and fix the sources behind any wrong answers.

A rename is a positioning decision, and AI assistants are now part of how buyers hear your positioning. Treat them as an audience in the launch plan, not an afterthought.

If you want help planning the AI search side of a rename, or auditing how assistants describe your product now, reach out. I am happy to take a look with you.