How to Check Product Names in AI Search Before a Rebrand Launch

What should you check before choosing a name?

How do you test a candidate name in AI assistants?

Run a small set of prompts for each candidate in ChatGPT, Gemini, Google's AI features, Perplexity and Copilot:

• "What is [Name]?"
• "Is [Name] a company or product?"
• "[Name] [category]", such as "[Name] payroll software".
• "Who makes [Name]?"
• "[Name] reviews"

Read the answers for three risks:

• An established entity. A confident description of another company or product means your name will be confused with it.
• A strong generic meaning. A common word may never become strongly associated with you.
• Category bleed. If assistants link the name to a different category, buyers searching for yours may not find you.

Repeat the tests in each language you sell in, since meanings and existing entities differ by market.

What makes a name risky for AI search?

• Sharing a name with a well-known brand, even in another industry.
• Common dictionary words without a distinctive twist.
• Near-misses of competitors' names, which invite mix-ups.
• Acronyms that already stand for many things.
• Names that suggest a different category, such as a finance-sounding name for a design tool.

A risky name is not impossible, but it needs more work to establish: consistent pairing with a category descriptor, more third-party coverage and patience. Avoiding confusion with similarly named companies is covered in how to stop AI from confusing your brand with a similar-named company.

What about renaming a single product or plan?

Product and plan renames cause the same confusion on a smaller scale, and they are easier to overlook. Assistants and review sites keep quoting old plan names and prices, and buyers ask about products that no longer exist under that name. Before renaming, test the new product name the same way as a company name. At launch, state the old name on the product page ("previously called [old name]"). Update pricing tables, documentation, integration listings and marketplace entries on the same day. For sub-brands, say clearly which company makes the product, so assistants do not treat it as a separate business. Keep a short page listing current and former product names, which gives assistants one place to resolve old references.

How do you prepare before launch?

• Draft the new pages with the new name paired with a plain category description and "formerly [old name]".
• Plan redirects from every old URL to its new equivalent.
• Prepare structured data. Google's Organization markup supports an alternateName property for other names the organisation is known by.
• List every profile to update: Business Profiles, review platforms, marketplaces, app stores, directories, partner pages and documentation.
• Prepare an announcement that states the old and new names clearly, so news coverage links them.
• Record a baseline of how assistants describe the old name, to compare after launch.

What should happen on launch day?

1. Publish the new site copy with the new name, a plain category description and "formerly [old name]" on the home and About pages.
2. Switch on redirects and check them.
3. Update structured data with the new name and the old name as an alternate.
4. Update profiles and listings, starting with those assistants cite most.
5. Publish the announcement and send it to journalists and partners.
6. Ask partners and marketplaces to update their pages.
7. Keep a page answering "Is [new name] the same as [old name]?" for at least a year.

How long will assistants use the old name?

Assistants that search the web usually reflect the new name within weeks, once key pages and listings are updated and recrawled. Answers from a model's memory can keep using the old name much longer. Research on model knowledge cutoffs has found that what a model knows often lags the date its makers state. Old articles, reviews and directory listings also keep the old name alive. So keep the old name visible on your site, as "formerly...", until assistants consistently use the new one.

How do you monitor the transition?

Track prompts for both names monthly: "What is [new name]?", "What is [old name]?", "[old name] alternatives" and category prompts. Record whether answers use the new name, link the two correctly and describe the product accurately. Watch for assistants treating the new name as a different, unknown company. Setting up alerts for changes is covered in how to set up alerts for brand mentions in AI responses.

What might an illustrative case look like?

Consider an illustrative HR software company planning to rename its product "Harbor". Pre-launch checks show assistants associate Harbor with a well-known software security project, a marina chain and several unrelated startups. In German answers, a different company appears entirely. The team switches to a more distinctive name that assistants do not associate with anything. At launch, the site says "[New name], formerly [old name], is HR software for companies with 50 to 500 employees". Redirects and alternateName markup are in place, and the company updates its profiles on two review platforms the same day. Six weeks later, search-enabled assistants use the new name and mention the old one. Some answers from memory still use the old name for several more months.

What do people ask about rebrands and AI search?

Should we keep the old name on our site?

Yes, as "formerly [old name]", for at least a year, so assistants and people can connect the two.

Will a rebrand reset our AI visibility?

Not if the transition is handled well. Redirects, consistent profiles and clear "formerly" statements carry most of the evidence across.

Do we need new reviews after a rebrand?

No, if review platforms update your existing profile to the new name. Avoid creating new profiles that split reviews.

How many candidate names should we test?

Test every shortlisted name. The checks take an hour or two per name and can prevent months of confusion, so it is worth testing even names the team already likes.