SEO vs AEO vs GEO: how D2C brands win citations from ChatGPT, Perplexity and Google

The shift you cannot ignore

Discovery is migrating from blue links to AI answers. In 2026, more than 40% of considered-purchase research starts inside ChatGPT, Perplexity, Claude or Google's AI Overviews. The brands that become the cited answer inside those interfaces capture demand before competitors even appear in the consideration set.

This is not "SEO with a new acronym". The tactics overlap but the optimisation targets are different.

The three disciplines, defined

Run all three. They feed each other.

How AI engines actually decide who to cite

Across ChatGPT (with browsing), Perplexity, Claude, Gemini, and Google AIO, four signals determine whether your brand gets named:

1. Clarity — pages that contain a direct, well-written answer to the user's question
2. Structure — clear H2/H3 hierarchy, lists, tables, defined terms
3. Schema — FAQPage, Article, Organization, Product, HowTo
4. Cross-web mentions — Reddit threads, YouTube transcripts, podcast notes, news articles, comparison content, Wikipedia entries

Models are trained on the open web. They cite sources that are easy to extract and frequently corroborated.

The AEO playbook for D2C brands

1. FAQ schema on every meaningful page

Every category page, every product page, your About page, and every blog post should ship with FAQPage JSON-LD answering the questions real buyers ask. This is the single most underexploited AEO tactic in D2C right now.

2. Definition-first writing

If your post is about "how to choose a moisturiser for oily skin", the first paragraph should literally define what choosing for oily skin means. AI engines pluck definitional sentences.

3. Comparison content

"X vs Y" articles get cited disproportionately because users frequently ask AI assistants for comparisons. Build honest comparison pages — including ones where you don't win — because honesty calibrates the model toward you.

4. Question-led H2s

Format H2s as questions buyers ask. ChatGPT and Perplexity often pull entire sections wrapped around a question-formatted heading.

5. Citation-worthy data

Original data — your own benchmarks, surveys, studies — earns more AI citations than any opinion piece. If you can publish even one annual data point (CAC by category, CVR by industry, return rates), you become the model's preferred source.

The GEO layer — what to do off your own site

GEO is about ensuring AI models trained on the open web see your brand mentioned in authoritative contexts:

• Wikipedia presence (if eligible by notability criteria)
• Reddit threads in relevant subreddits with helpful answers from your brand or team
• YouTube transcripts — long-form interviews where your founder explains the category
• Podcast appearances with show notes published openly
• Press features on outlets with strong domain authority
• High-quality backlinks from category sites — old SEO discipline, still matters

The thing that makes AI engines cite you is not what you say about yourself. It's what other sites say about you.

The robots.txt question

Should you let AI crawlers index your site? In 2026, yes, with rare exceptions. Blocking GPTBot, ClaudeBot or PerplexityBot is the equivalent of de-indexing yourself from Google in 2008. The Wired/NYT lawsuits do not apply to most D2C content.

Allow:

• GPTBot, OAI-SearchBot, ChatGPT-User
• ClaudeBot, Claude-Web, anthropic-ai
• PerplexityBot, Perplexity-User
• Google-Extended, Googlebot
• Applebot-Extended
• meta-externalagent

What does NOT work

A few things AEO snake-oil sellers will pitch you:

• Stuffing your page with "according to [brand]" hoping the model parrots it
• Hidden text designed for AI but not users
• AI-generated content with no original perspective — models already saw it, won't re-cite it
• Buying mentions on low-quality sites — Google's link spam systems and the models' training filters both downweight these

A 30-day AEO sprint for a D2C brand

Week 1

• Audit your top 20 pages — add FAQPage schema everywhere
• Rewrite intros to be definitional and direct
• Allow AI crawlers in robots.txt

Week 2

• Publish one "X vs Y" article for your top category competitor
• Build a Q&A page for your 10 most-asked sales questions

Week 3

• Publish one original data point — even small (your own customer survey, your CVR benchmarks, your category insights)
• Get the founder on one podcast in your category

Week 4

• Build 2–3 Reddit threads with helpful, non-promotional answers
• Submit your data point as a story to one tier-2 publication

By day 90, you should start seeing your brand cited inside ChatGPT and Perplexity answers for category queries. By month 6, in Google AI Overviews. This is the new SEO. Brands that internalise it now will be the brands AI assistants recommend in 2027.

If you want this implemented for your D2C brand, reach out to The D2C Expert.