What is answer engine optimisation? Honest definition.

AEO, GEO, LLMO - which is which

The acronym soup confuses everyone. Here is the practical sorting:

Term	Origin	Audience that uses it
AEO	Voice search era (in use by February 2018), expanded for AI chatbots	Traditional SEO leaders, agencies, marketing teams from a search background
GEO	November 2023 academic paper by Princeton and IIT Delhi researchers, accepted to KDD 2024	Younger marketing leaders, SaaS founders, AI-literate teams
LLMO	Practitioner term, in use by October 2023	Developer-led teams, technical marketing, devrel

In delivery the work overlaps almost entirely. The same schema work, the same content depth work, the same named-author work. Pick whichever term your audience recognises.

Google now takes the same line for its own search. Its guide to optimising for its AI features, published in May 2026 and last updated on 10 July 2026, names AEO and GEO and says optimising for generative AI search "is optimizing for the search experience, and thus still SEO". It also lists so-called hacks to ignore for Google Search: llms.txt files, chunking content, rewriting content just for AI, seeking inauthentic mentions and overfocusing on structured data. That covers Google only. ChatGPT, Perplexity and Claude run their own systems.

What AEO involves

• FAQ schema. AEO traditionally leaned heavily on FAQPage schema, a habit from the voice-search era. Google stopped showing FAQ rich results on 7 May 2026, and its AI guide says there is no special schema.org markup you need for its AI features. Write the questions and answers as visible text first; I no longer treat the markup as a citation lever.
• Direct-answer paragraphs. One-sentence or two-sentence answers at the top of each section. The shape voice assistants and AI engines pull cleanly.
• Question-led page architecture. Pages structured around the questions buyers actually ask, with each question answered explicitly.
• Conversational query forms. Targeting "how do I", "what is the best", "why does my", "should I" - the queries voice and AI users actually phrase.
• Schema for products, services, organisations. The structural credibility layer.

How AEO and GEO differ in practice

Not much. The honest list of differences in 2026:

• AEO practitioners emphasise FAQ schema and conversational query forms more (legacy of voice search).
• GEO practitioners emphasise citation tracking and Reddit/community presence more (legacy of the academic paper and the SaaS adoption).
• AEO often skips the AI bot access work that GEO foregrounds. The crawler that matters for ChatGPT search is OAI-SearchBot: OpenAI says sites that opt out of it will not be shown in ChatGPT search answers, though they can still appear as navigational links. GPTBot is the crawler for model training, and each setting is separate.

If you are buying a service, ignore the acronym and ask what the deliverable actually includes. The acronym is a tell, not a contract.

AEO vs SEO: when each one applies

SEO targets the ranked list of blue links Google has served for 25 years. AEO targets the direct answer Google AI Overviews, AI Mode, ChatGPT, Perplexity and Claude now serve above (or instead of) those blue links. By Google's own count, AI Overviews has over 2.5 billion monthly active users and AI Mode more than one billion monthly users. Two surfaces, two scoring weights, mostly overlapping work.

• Where SEO is still primary: commercial-intent queries (buying a product, comparing vendors, finding a service near you). Click still happens. Rank still matters. In Seer Interactive's April 2026 study of 53 brands, AI Overviews showed on 8% of commercial and 5% of transactional queries, against 36% of informational ones.
• Where AEO is now primary: definitional queries ("what is X"), how-to queries ("how do I X"), comparison framing queries ("X vs Y") - all of which get a direct AI answer above the SERP. In the same Seer sample, comparison queries triggered an AI Overview 95.4% of the time and question-style queries 85.9%. Clicks fell hard wherever AI summaries showed: Pew found people clicked a result on 8% of visits with an AI summary against 15% without (March 2025, 900 US adults), and Ahrefs measured a 58% lower click rate for the top-ranking page when an AI Overview shows (December 2025 data, 300,000 keywords). Seer then saw organic click-through on AI Overview queries partly rebound, from 1.3% in December 2025 to 2.4% in February 2026, still below the 3.8% on queries without an AI Overview.
• Where both apply equally: branded queries, navigational queries, transactional product pages. The click and the citation are both available.

For most UK B2B brands in 2026 the right move is to budget AEO and SEO as one programme. My rough rule of thumb is about 70% shared work and 30% AEO-specific. Splitting them between two consultants double-pays for one job.

How to measure AEO success

SEO measures rank position and organic click. AEO measures three different things:

1. Citation frequency. How often your brand appears as a named source inside AI engine answers on your priority queries. Sample weekly with a tool (Profound, AIClicks, Otterly) or manually through ChatGPT/Perplexity browser windows.
2. Citation share-of-voice. Your appearances as a percentage of total citations against named competitors on the same queries. Useful for board reporting.
3. Query coverage. How many of your priority queries cite you at all (binary present/absent across the priority list).

Two free first-party reports now sit alongside those tools. Google Search Console has a Generative AI performance report showing impressions, which pages appear in AI responses and in which countries; it reached all websites worldwide by 31 August 2026. Search Console also has a new control to opt out of AI Overviews and AI Mode, and Google says sites that opt out will not receive traffic or impressions from its generative AI features. Bing Webmaster Tools has had an AI Performance report in public preview since February 2026, counting citations across Microsoft Copilot and Bing's AI summaries, with Intents, Topics, Citation Share and Compare added in preview in June 2026. Google's AI guide also warns that no third-party tool has access to its internal ranking or AI systems, so treat tool scores as samples, not the full picture.

What about an AEO grader? HubSpot's AI Search Grader, at hubspot.com/ai-search-grader, is a free, one-time check of what ChatGPT, Perplexity and Gemini say about your brand, scored out of 100, with no account required. The old /aeo-grader address redirects there. A paid version costs $50 a month, with no HubSpot subscription needed. It is a useful snapshot for a first conversation. It does not test Google AI Overviews, AI Mode, Copilot or Claude, so it cannot replace the reports above.

What to ignore: raw traffic from AI engines in GA4 alone. AI engines often cite without sending a click - your brand mention is the win, not the visit. Use GA4 as a directional secondary metric, not the primary one. Reporting cadence: weekly or fortnightly. Daily creates noise (AI outputs vary per session). Monthly is too slow if you are actively running work.

The honest position

For most UK buyers in 2026 the right answer is: get a senior SEO consultant who has matured into AI search work. The label they use (SEO, AEO, GEO, LLMO) matters less than whether they have actually run citation tracking, checked that AI search crawlers such as OAI-SearchBot can reach the site, written question-and-answer blocks and seen citations land. Watch for anyone who sells Google-Extended as an AI Overviews fix: Google says it does not affect a site's inclusion in Google Search, and that it covers Gemini model training and grounding in Gemini Apps and Vertex AI. Pick the practitioner, not the acronym.