Cannabis & CBD
Someone asked ChatGPT a question about your category, and a competitor's name came up. Why did their name come up but not yours? The question you're asking now is the right one. You need to understand what ChatGPT actually looks at when it decides which brand to name.
Key Takeaways
ChatGPT assembles a composite picture from everything it can read about a topic across the open web. When it needs to name a brand, it picks the one it can describe confidently, with claims that hold up across multiple independent sources. The brands that get mentioned tend to be the ones whose identity, content, and third-party presence all tell the same story.
ChatGPT evaluates three signals when deciding which brand to mention. In search-grounded modes it pulls from live web results; otherwise it draws on its training data. Either way, the same three things determine who gets named.
1. Entity identity. ChatGPT needs to know exactly who your brand is, what you do, and where you operate. If your site, directory listings, and press mentions each describe your name, category, and service area differently, AI sees noise. A study of 75,000 brands by Ahrefs found that web mentions correlate 0.664 with AI brand visibility, roughly three times more than backlinks at 0.218. Where your name appears and whether the description matches matters more than how many links point to your site.
2. Answer-first content. A citation study by Kime.ai found that 44% of ChatGPT's citations come from the first third of the page. If the answer to the question appears in your opening sentences, AI is far more likely to cite you. The Princeton GEO study found a similar pattern: adding concrete statistics boosted AI citation rates up to 40%, and adding source citations boosted them up to 41%. Keyword density had no measurable effect.
3. Third-party corroboration. Reviews, directory listings, press mentions, industry profiles. If the only place that describes you as the best dispensary in your market is your own About page, AI has no independent reason to repeat it. If Leafly, a local news outlet, and a cannabis directory all describe you the same way, AI has three reasons to trust that description.
Social follower counts, ad spend, and keyword stuffing have no meaningful effect on AI visibility. A Semrush study of 50,000 brands tested whether traditional SEO metrics predict who ChatGPT names. The brand with the highest Domain Authority in a category was the topic owner only about half the time. The brand with the most organic traffic did no better. Meaning, traditional SEO strength alone does not determine who AI will mention.
Brands get left out when AI cannot read their information, cannot distinguish them from other entities, or cannot find enough independent sources to trust. Four factors explain:
1. Trapped information. Menu widgets, interactive maps, login-gated directories, PDFs, images of text. Your data exists, but ChatGPT cannot extract or quote it. The Kime.ai citation study found that AI extracts information from tables at an 81% rate, compared to 23% for identical information written as paragraphs. If your best content lives in a format AI cannot parse, you are invisible regardless of how good it is.
2. Name collisions. When multiple businesses share a name, or when a brand name doubles as a common word, AI hedges. It picks the wrong entity or qualifies its answer so heavily that no one gets a clear mention. Distinctive names and consistent category language across every source eliminate the ambiguity.
3. Stale content. Pages updated within 30 days get 3.2 times more AI citations than pages older than 90 days. AI trusts recent information over old information, and content you published a year ago and never touched is working against you.
4. Thin presence across prompts. The Semrush study found that a brand needs to appear in four out of five related prompts to own a topic. Brands with a five-point lead retained first place 90% of the time month over month. A single strong page is not enough. AI looks for consistency across the web.
Eighty-five percent of ChatGPT categories have no clear brand owner. The space is wide open.
Video is the strongest correlated signal for AI brand visibility. That same Ahrefs study found that YouTube mentions correlate 0.737 with AI visibility, higher than backlinks, domain authority, or content volume.
AI transcribes YouTube videos and indexes them as text. A short film published alongside a written piece gives AI a second independent source describing the same information, doubling the surface area AI can cross-reference. Production value is secondary to having the information in video form, because AI reads the transcript, not the cinematography. Cannabis brands already building AEO programs with video stack video into the workflow with every piece they publish.
Yes. Cannabis brands already invest heavily in organic marketing: SEO, directories, Weedmaps, Leafly. The tactics that answer engine optimization rewards are different: entity identity structured for machine comprehension, answer-first content, and third-party corroboration AI can verify. Being strong in traditional organic does not automatically make you visible to AI, and almost no cannabis brands have started building these signals yet.
Three steps build AI visibility from where most cannabis brands are today:
1. Align your entity identity. State your brand name, category, and service area the same way on your site, your directory profiles, your press mentions, and your YouTube channel. AI needs a consistent description it can repeat without contradicting itself.
2. Publish answer-first content with video. Answer the questions your buyers actually ask, with the answer in the first two sentences and real specifics backing it up. Publish a video version alongside every page to give AI a second independent source to cross-reference.
3. Earn third-party mentions. Get described in places AI reads independently: reviews, directories, press, industry publications. The more independent sources that describe your brand the same way, the more confidently AI names you.
In most industries, paid media would be a fourth step. Cannabis has real paid channels (endemic programmatic, Weedmaps and Leafly placements), but the easy self-serve paths like Google Ads and Meta are blocked, so fewer brands are competing for organic visibility. AI favors established pages, and every month of signal-building makes it harder for a competitor to close the gap. The brands that start now will be the ones AI describes confidently when your buyers ask. AVO Brands, the boutique creative agency for regulated consumer products: cannabis, beverage, and CPG, builds these programs from a measured baseline, with answer-first content and short films on every piece.
Do people actually use ChatGPT to find cannabis products? Yes. Buyers ask AI for dispensary recommendations, product comparisons, and category questions the same way they ask about restaurants or electronics. As AI tools become more routine, the brands AI can confidently describe are the ones that get named.
Does ChatGPT use my Google Business Profile? Partly, but if your hours, menu, and service area only exist on a map listing or inside a widget, ChatGPT cannot extract them. It needs the information in plain text on a page it can read.
How long until ChatGPT picks up a new page? There is no fixed timeline. AI favors established pages, so starting sooner means compounding sooner. AI Overviews rotate faster than ChatGPT's core recommendations.
Can I pay to be mentioned in ChatGPT? Not yet. OpenAI is building toward an ad product. Paid placements are already live on Perplexity, Bing Copilot, and Google AI Overviews. Even where paid exists, organic signals still drive the core answers.
