Brand situation

You Are Launching a New Brand and AI Has Never Heard of It

A brand that launched after an AI model's training cutoff does not exist in that model's built-in knowledge, so it can only be found through live search. That means your first months are about getting indexed, getting described consistently on real third-party publications and checking that AI crawlers can reach your site.

5 min readPublished October 14, 2026By the Odys Blue Ocean team

If your brand launched recently, most AI models have no memory of it at all. Anthropic’s newest Claude models list a training data cutoff of June 2026, and OpenAI’s current API models list knowledge cutoffs in spring 2026. Anything that happened after those dates, including your launch, reaches an answer only when the engine searches the live web and finds pages that describe you. Your job in the first months is to make sure those pages exist, are indexed and say the same thing.

What AI likely says about you today

Ask ChatGPT, Gemini or Perplexity a few direct questions and you will typically see one of three results. For “what is [your brand]?” the answer is either a polite “I don’t have information about that”, a guess based on the words in your name, or a confident description of a different company with a similar name. For category prompts such as “best new [category] brands in 2026” or “alternatives to [the market leader]”, you are usually absent while established names fill the list.

That is normal for a new brand, not a verdict on your product. In a study of 75,000 brands, 26% had zero AI Overview mentions, and that sample was not limited to new brands. Before you change anything, write down ten prompts a real buyer would type, run each one a few times in several engines and save the answers. The free AI visibility self-check generates a starter prompt set and a scoring grid for exactly this.

What the engines look for

A new brand depends on retrieval-augmented generation: the engine runs a search, reads a handful of pages and writes an answer from them. Three things decide whether you are in that handful.

Studies point the same way. Across those 75,000 brands, how often a brand was mentioned on the web had the strongest correlation with AI Overview visibility (0.664). That is correlation, not proof that mentions cause visibility, but it matches how retrieval works: the more relevant pages describe you, the more chances you have to be read.

Three risks specific to this situation

1. A name collision. New brands often share a name, or most of one, with an older company, a product line or a place. With no data about you, an engine may describe the older entity and present it as you, a kind of hallucination. Use your full brand name plus category in every early mention (“Pacewell, the Austin running shoe brand”, to take an invented example), and add sameAs links so systems can separate the two.

2. Crawlers still blocked from staging. Many sites launch with the robots.txt file from their staging environment, which disallows every bot. Check yours on launch day. Blocking GPTBot only opts you out of training, but blocking OAI-SearchBot, PerplexityBot or Googlebot can keep you out of the live answers you need most. OpenAI says robots.txt changes take about 24 hours to reach its systems.

3. One loud week, then silence. A launch often produces a burst of coverage that stops after a week. Retrieval can only use what is still indexed and relevant, and the sources engines cite shift: Semrush saw Reddit go from appearing in about 60% of ChatGPT responses to about 10% within weeks in 2025. Plan mentions that keep arriving after launch, not one spike.

What to do this month

  1. Run your baseline: ten buyer prompts, three runs each, in ChatGPT, Gemini, Perplexity and Google. Save the answers with dates.
  2. Check robots.txt and make sure OAI-SearchBot, PerplexityBot, Claude-SearchBot and Googlebot are allowed. Submit your sitemap in Google Search Console and Bing Webmaster Tools.
  3. Write a one-paragraph brand fact sheet: name, category, what you sell, who it is for, where you operate, founding year. Use the same wording on your site, LinkedIn, social profiles and press kit.
  4. Add Organization schema with your logo and sameAs links on the home page or about page.
  5. Get your first independent mentions live in publications that cover your category, each describing you in context (category, use case, market).
  6. Re-run the same prompts at the end of the month and compare.

Your options

Option What it means for a new brand
Do it yourself Fix your site and facts, pitch founders’ stories to niche publications; slow but builds skills
Digital PR Strong for launch news; journalists cover what is newsworthy, so it rarely continues month after month
Brand mentions Placed in context on real, indexed publications; steady, controllable, and you see every live URL
Wait Future models may learn about you from whatever the web says by their cutoff, which may be little

With The Blue Ocean Brand Mentions, first mentions typically go live within 30 minutes of kickoff, with more added over time. Live means published, not cited: when AI engines pick them up depends on indexing, retrieval and model updates, and it varies by engine and topic. That is why we measure before and after instead of promising a result. Our comparison of brand mentions and digital PR explains how the two fit together at launch.

Every placement should be real editorial content written for that publication’s readers, follow the publisher’s rules and the disclosure rules of each market, and carry rel=”sponsored” or rel=”nofollow” on any paid link. (This is not legal advice; disclosure rules differ by market.) The longer playbook is in how a new brand gets into AI answers.

What to do next

  • Fix crawler access and publish your brand fact sheet before anything else.
  • Record a dated baseline so you can tell later what actually changed.
  • Get a free AI visibility check and see where your new brand stands across the engines today.
  • Later, once AI knows the new brand, growing how high ChatGPT recommends it is the job of The Blue Ocean GPT.

Frequently asked questions

Why does ChatGPT say it has never heard of my new brand?

Because your brand launched after the model's training data was collected, and the answer was given without a live web search, or the search found nothing useful. Models only know what was in their training data up to a cutoff date. Anything newer has to be retrieved from indexed pages, so the fix is real, consistent coverage that search systems can find.

How long until AI mentions a new brand?

Nobody can honestly give a date. Retrieval-based answers can pick up a brand once indexed pages describe it, while built-in model knowledge only changes when a new model with a later cutoff ships. Answers also vary from run to run, so measure how often you appear across many prompts over several weeks rather than waiting for one perfect answer.

Should I wait until after launch to think about AI visibility?

No. The cheapest time to fix crawler access, structured data and consistent brand facts is before launch, while the site is still being built. Launch-week coverage also matters more if your own site, profiles and early mentions all describe the brand the same way, so the engines see one clear story from day one.

Can I just create a Wikipedia page for my new brand?

Usually not, and you should not write it yourself. Wikipedia asks people with a conflict of interest not to edit articles about themselves and requires paid editors to disclose who pays them. A new brand rarely has the independent coverage Wikipedia expects. Wikidata has lower notability rules, but it still needs serious public references.

Your next move

See what AI says about your brand right now.

Tell us your website, your topic and your market in about a minute. We run a free AI visibility check and show you where you stand.

Get my free AI visibility check →
About 60 secondsFirst mentions live in 30 minutesEvery placement reported
Free AI visibility check →