AEO for Cannabis Brands: Citation Strategy in a Regulated Vertical

Why cannabis brands need AEO more than any other vertical

Cannabis sits in a uniquely punishing acquisition landscape. Google Ads blocks the category outright. Meta blocks the category outright. TikTok blocks the category outright. Most programmatic networks block the category outright. The few platforms that allow paid placement charge punitive CPMs and limit creative.

That leaves three real channels: organic search, dispensary partnerships, and word of mouth. Now there is a fourth, and it is growing faster than the other three combined. AI assistants get asked about cannabis constantly. Strain recommendations. Dosing questions. Legal status in specific states. Edibles versus flower. Terpene profiles. CBD versus THC. The answers users get back shape brand preference before they ever open a dispensary menu.

Cannabis brands that earn citations in these answers compound visibility every month. Brands that do not are invisible to the channel entirely. There is no middle ground because AI answers tend to mention two to four sources and skip everything else.

How AI models currently answer cannabis queries

We pulled 300 cannabis prompts across ChatGPT, Claude, Gemini, Perplexity, and DeepSeek in May 2026. The citation pattern is consistent across models.

The 9 percent share that brand-owned hubs capture is the opening. Most cannabis brands publish either no education content or thin marketing-flavored posts. The brands that publish genuinely useful, structured content move into the citation set quickly because the competitive field is shallow.

The content categories that earn cannabis citations

Five content categories drive almost every cannabis citation we have measured. Build these out before anything else.

Strain and product education comes first. Every cited cannabis page we analyzed includes structured data on cannabinoid profile, terpene breakdown, effects, and use cases. AI models cite pages that answer the user's actual question (which terpene profile helps with sleep, what THC percentage is appropriate for a beginner) rather than pages that simply list SKUs.

State-by-state legal pages come second. Cannabis users ask AI assistants about legality constantly. Brands that publish current, accurate state guides become the cited authority for those queries, even when they do not operate in that state. Maintenance matters here. Stale pages drop out of citations within 60 to 90 days.

Dosing and consumption guides come third. Edibles dosing, tincture dosing, microdosing, first-time use, tolerance reset. These queries are high intent and low competition because most brands avoid them out of legal caution. Brands willing to publish responsible, clearly worded guidance with appropriate disclaimers capture the category.

Method comparisons come fourth. Flower versus vape, edibles versus tincture, indica versus sativa, full-spectrum versus isolate. Comparison content earns citations at roughly 2.5 times the rate of single-topic content because AI assistants are often asked comparative questions directly.

Dispensary locator and menu pages come fifth. Local cannabis queries are massive. Brands that publish clean, schema-rich dispensary pages get cited for local "where to buy" questions even when the user did not name the brand.

Compliance language that does not kill citations

Cannabis content has to navigate state and federal compliance. The mistake most brands make is layering so much disclaimer language that the substantive content gets buried. AI models rank content by answer density, and pages that take 800 words to deliver a 60-word answer lose to pages that deliver the answer in the first paragraph.

The pattern that works: lead with the answer, follow with structured supporting detail, close with appropriate medical and legal disclaimers. Keep disclaimers in a consistent footer block so they do not interrupt the body. Use plain language for the substantive content. Avoid coded language like "wellness" or "botanical alternative" when the user is clearly asking about cannabis. AI models penalize evasive content.

Schema and technical setup for cannabis sites

Cannabis sites should deploy four schema types at minimum. Article schema on every education page. Product schema on every SKU page (cannabinoid profile, terpene profile, weight, strain type). FAQ schema on dosing and education pages. LocalBusiness schema on every dispensary location page.

Beyond schema, three technical points matter more than usual in this vertical. First, robots.txt should explicitly allow GPTBot, ClaudeBot, PerplexityBot, and Google-Extended. Many cannabis sites block these by default through cautious agency setup. Second, image alt text should describe the actual product or scene rather than generic marketing copy. Third, structured FAQ blocks should appear above the fold on education pages because AI crawlers preferentially extract from the visible portion of the page.

What we have seen work, with timelines

A multi-state operator we worked with in late 2025 launched 60 strain education pages and 14 state legal guides over 90 days. By month three, they were cited in 22 percent of generic cannabis queries we tracked across ChatGPT and Perplexity. By month six, citation rate held above 30 percent and they were the most-cited brand in the category for "best strain for sleep" and "first-time edibles" queries.

A single-state dispensary brand in the Northeast deployed 40 dispensary location pages, each with LocalBusiness schema, hours, menu data, and a local education FAQ. Within 60 days they were cited in 41 percent of "dispensary near me" queries inside their service area. The competitive moat in local cannabis is small because most dispensaries publish poorly structured pages.

The pattern holds. Cannabis brands that publish structured, useful content earn citations faster than brands in any other vertical we measure because the competitive field is shallow and AI assistants are actively looking for credible sources.

The 12-month plan for a cannabis brand starting from zero

Months one and two: publish 30 strain education pages and 10 state legal guides. Deploy Article, Product, and FAQ schema. Audit robots.txt and unblock AI crawlers. Set up monthly citation tracking against a defined prompt set.

Months three and four: publish 20 dosing and consumption guides plus 10 method comparison pages. Begin dispensary locator buildout if applicable. Refresh state guides quarterly.

Months five through eight: expand into long-tail education (specific terpenes, cannabinoid minors like CBG and CBN, condition-specific guides where compliance allows). Begin earning third-party mentions in cannabis trade media. Most citation lift in this window comes from off-domain authority signals.

Months nine through twelve: optimize for the queries you have not yet captured. Build the comparison and "best of" content that competitors are using to outrank you. Most brands plateau at 25 to 35 percent visibility around month nine without this push.