Generative engine optimization (GEO)

Generative engines answer a question in their own words instead of returning a list of links. Examples include ChatGPT, Perplexity, Gemini, Microsoft Copilot and Google's AI Overviews. GEO is the practice of making sure that, when a customer asks one of them about your kind of business, the answer can include you, with the right details.

The term comes from a research paper, "GEO: Generative Engine Optimization", first posted in November 2023 and accepted to the KDD 2024 conference. Its authors describe generative engines as systems that "use generative models to gather and summarize information to answer user queries", typically by synthesizing information from multiple sources.

How it works

GEO builds on SEO. An AI answer that draws on web pages can only use pages it can find and read, much like retrieval-augmented generation. Google says the best practices for SEO remain relevant for its AI features, and that no special files, markup or optimizations are needed to appear in them. The practical work:

• State your facts the same way everywhere. Name, services, areas served and contact details should match on your site, business profiles and directories, so a model doesn't find conflicting versions.
• Answer questions in clear, quotable sentences, with the answer first.
• Earn mentions on other reputable sites, since answers often combine several sources.
• Measure it. Ask a fixed set of customer questions of each assistant every month, and record whether you're mentioned and whether the details are right.

Nobody controls what an AI assistant says. Honest GEO work commits to effort and measurement, not to mentions.

How it relates to Worcia

GEO is half of SEO & GEO, Worcia's search and AI-answer optimization service, because the two share most of the work. For the longer guide, read What is GEO?