GEO: getting your brand into ChatGPT's answers
Generative engines don't rank links: they compose an answer and cite their sources. A complete method to exist in ChatGPT, Perplexity and Gemini answers — before the positions freeze.
Your clients don't "google" anymore, they ask. ChatGPT, Perplexity, Gemini. And when the engine answers, it cites three or four sources — not ten blue links. If your brand isn't there, you don't exist. That's the whole point of GEO (Generative Engine Optimization), and it's the workstream I industrialized with GetSpotted.
LLMs don't rank, they cite
Classic SEO optimizes for a ranking. Generative engines compose an answer and lean on sources they deem reliable and citable: structured pages, dated and quantified data, clear entities, identifiable authors. A vague "corporate" page doesn't stand a chance; a page that precisely answers a question with sourced numbers does.
The method, in 4 steps
1. Analysis: query the engines on the requests that matter to you and map who gets cited, from which sources. 2. Sourcing: identify the media and pages LLMs already cite in your vertical. 3. Production: create citable content — direct answers, data, structured FAQs, schema.org. 4. Tracking: measure your share of voice in generative answers, month after month.
What it changes in practice
At Linkar, my insurtech, acquisition is 100% organic — zero ad budget. Traffic from AI engines often converts better than classic SEO: the user arrives pre-qualified, the engine's answer has already done the reassurance work. GEO is not a monitoring gadget: it's a fully-fledged acquisition channel, still barely contested.
The right time to start is now: LLM citation patterns freeze fast, and brands installed in the answers will be hard to dislodge.