Short answer: AI assistants cite sources they can understand and trust. Write pages that answer buyers’ questions directly, mark them up with structured data, keep your company facts consistent everywhere, and earn mentions on other credible sites.
More B2B buyers now start research by asking an assistant instead of typing into a search box. “Which agencies in Riyadh handle government tenders?” “What should an ERP implementation cost in Egypt?” The answer names a handful of companies and links to a few sources. If yours is not among them, you are not on the shortlist.
Two terms have appeared for the work of getting there. Generative engine optimisation (GEO) is about being used and cited in AI-generated answers. Answer engine optimisation (AEO) is about being the direct answer to a question, whether in an assistant or in a search engine’s answer box. In practice they overlap, and most of the work builds on good SEO.
How assistants choose what to cite
The exact methods differ by product and change often, but the broad pattern is stable. Many assistants search the web, read a set of pages and write an answer from them. Pages that state a clear answer near the top, come from a site the system treats as credible and agree with other sources are more likely to be used.
Write answer-first pages
For each important question your buyers ask, have a page or a section that answers it in the first two or three sentences, then gives the detail. Use the question, or something close to it, as a heading. Short definitions, numbered steps and clear comparisons are easy for both people and machines to lift.
Service pages benefit from the same approach. State plainly what the service is, who it is for, where you offer it and how it is priced or scoped. Vague copy written to sound impressive gives an assistant nothing to quote.
Add structured data
Structured data, written in schema.org vocabulary as JSON-LD, describes your pages in a form machines read reliably. For a B2B company the useful types are usually Organization, Service, FAQPage, Article and BreadcrumbList. Mark up only what is visible on the page, and keep it accurate.
Keep your company facts consistent
Assistants build a picture of your company from many places: your website, LinkedIn, business directories, press coverage and partner sites. If your services, locations or descriptions differ between them, the picture blurs. Write one short description of the company and use it everywhere, and keep your name, address and contact details identical.
Earn mentions elsewhere
Being mentioned on other credible sites matters as much for AI search as it does for classic SEO. Useful sources include industry associations, partner directories, client case studies published on the client’s own site, interviews and articles in trade media. Each mention should describe what you do in the same terms you use yourself.
Make the site easy to read
- Pages that load quickly and render their text without heavy scripts
- A clean heading structure, one H1 per page
- An XML sitemap submitted to Google Search Console and Bing Webmaster Tools
- A robots.txt that does not block the crawlers you want to be read by
- Optionally, an llms.txt file that summarises the company and links key pages. It is a proposed convention, so treat it as a small extra rather than a fix
Measure it
Measurement is still rough. Track referral traffic from assistants in your analytics, ask new leads how they found you, and run a fixed list of buyer questions through the main assistants every month to see whether you are named. Over a few months, the trend is more useful than any single answer.
Frequently asked questions
What is GEO?
Generative engine optimisation (GEO) is the work of making a website more likely to be used and cited in answers written by AI assistants such as ChatGPT, Gemini and Perplexity.
What is the difference between GEO, AEO and SEO?
SEO aims to rank pages in search results. AEO aims to provide the direct answer to a question. GEO aims to be cited in AI-generated answers. They share most of the same groundwork: clear content, structured data and a credible site.
Does llms.txt help with AI search?
llms.txt is a proposed convention for summarising a site for language models. Support varies, so it is a low-cost extra, not a replacement for clear content and structured data.



