When a buyer asks an AI assistant “what is the best time tracking tool for a small agency”, the answer names a handful of products and sometimes links to sources. Generative engine optimisation, or GEO, is the work of being one of the names and one of the sources.
The mechanics differ by product, but they share a pattern you can work with.
Short on time? Comppound does this work for SaaS companies. AI-native, and a person reviews every page.
Have us do itTwo places an answer comes from
An assistant answers from what the model learned in training and, increasingly, from pages it fetches at the moment you ask. The first is slow to change and hard to influence directly. The second, often called retrieval or grounding, uses a search index and your live pages, and you can influence it in the same way you influence search.
What retrieval looks for
When an assistant searches the web to answer, it tends to favour pages that:
- rank or are indexed for the searches it runs behind the scenes
- its crawler is allowed to read, without needing to run JavaScript
- state an answer plainly in a passage it can lift
- are corroborated by other sources, such as reviews and community threads
Comppound can do this for you
Want to be named in AI answers?
Comppound checks what ChatGPT, Perplexity, Gemini and Google's AI features say about your category, opens your site to their crawlers and builds the pages and mentions they cite. AI-native, with every page reviewed by a person.
- AI visibility baseline for your buyer questions
- Crawler access and quotable pages
- Monthly tracking of where you are named
Why one question becomes many searches
Assistants often break a question into several searches, a step called query fan-out. “Best time tracking for agencies” might become searches for alternatives to a market leader, pricing, integrations and reviews. A site with good pages for each of those has more chances to be picked up.
What you control
You control whether crawlers can read you, how quotable your pages are, how complete your coverage of buyer questions is, and partly how often you are mentioned elsewhere. The rest of this track takes those one at a time.
Worked example · an illustration, not a client result
Tracing one AI answer
Tallyroom asked three assistants the same buyer question and wrote down every source each one named. Most sources were comparison and alternatives pages, plus one community thread. None was Tallyroom's own site, because it had no page of either kind.
| Source type named | Could Tallyroom influence it? | How |
|---|---|---|
| Review-site list | Partly | Ask customers for reviews |
| Competitor's alternatives page | No | Publish a fairer one of its own |
| Community thread | Partly | Take part honestly where it is relevant |
| Independent blog comparison | Partly | Offer accurate product details to the author |
Do it yourself
Ask the assistants yourself
- 1Write five questions a buyer would ask before choosing a product like yours.
- 2Put each to ChatGPT, Perplexity and Gemini, and note whether Google shows an AI Overview.
- 3Record every product named and every source linked.
- 4Mark where you are named, where a competitor is named instead and where nobody is.
- 5Keep the sheet. It is your baseline for the tracking lesson.
Free tool for this: Free AI visibility report, AI crawler access checker
Common mistakes
Treating AI search as separate from SEO
Retrieval leans on search indexes and crawlable pages. Most GEO work is good SEO done with quoting in mind.
Judging by one answer
Answers vary from run to run and between users. Ask several times and look for patterns.
Chasing a trick
There is no hidden tag that makes an assistant name you. Coverage, clarity and corroboration do.
Rather not do it yourself?
Want to be named in AI answers?
Comppound checks what ChatGPT, Perplexity, Gemini and Google's AI features say about your category, opens your site to their crawlers and builds the pages and mentions they cite. AI-native, with every page reviewed by a person.
- AI visibility baseline for your buyer questions
- Crawler access and quotable pages
- Monthly tracking of where you are named
Questions
They are different names for much the same work: being cited or named in AI-generated answers.
Some AI products sell ads beside answers, but the answer itself is not for sale. Being named comes from the sources it trusts.
Key terms
- Generative Engine OptimizationGEO is the work of making a brand and its pages easy for AI search tools like ChatGPT, Perplexity and Google AI Overviews to find, understand and cite.
- Retrieval-Augmented GenerationRAG is a method where an AI model first retrieves relevant documents, then writes its answer using them, instead of relying only on what it learned in training.
- GroundingGrounding is tying an AI model's answer to specific, verifiable sources, such as search results or documents, so it relies less on memory alone.
- Query Fan-OutQuery fan-out is when an AI search system splits one question into many related searches, runs them together and combines the results into one answer.
- AI CitationAn AI citation is a link or source reference that an AI search tool shows beside its answer to point to the page it drew on.
- AI OverviewsAI Overviews are Google's AI-written summaries shown at the top of some search results, with links to the web pages used to build them.
- Parametric KnowledgeParametric knowledge is what a model learned during training and stores in its weights, as opposed to information it retrieves at the time of answering.
