Generative engine optimization (GEO) for founders: what it is and what to do first

You'll see several labels for overlapping work. They're mostly vocabulary, so don't let them slow you down.

Pick one term for your team and move on. The tactics are the same.

Much of GEO is good SEO with a different scoreboard. Google's guidance on AI features and your website (opens in a new tab) says the usual SEO best practices apply to AI Overviews and AI Mode, and that there are no additional technical requirements beyond being indexed and eligible to appear in Search with a snippet.

The shared foundations:

• Pages that can be crawled and indexed, with the content present in the HTML.
• Titles, headings and descriptions that say plainly what each page is about.
• Content written for people, with specifics and first-hand experience.
• Links and mentions from other reputable sites.
• Structured data that describes your organization and products.

If your SEO basics are broken, fix those first. Google's AI answers are built on its search index, and the other assistants lean on web search too.

The overlap is real, but five differences change what you prioritize:

1. The win is a mention. The answer often satisfies the buyer without a visit, so what matters is whether you're named, how you're described and whether you're linked.
2. Third-party pages carry more weight. AI answers synthesize many sources. A "best tools" article that names three competitors and not you can cost you more than a missing post on your own blog.
3. Entity clarity matters. The model has to understand what your company is and which category it belongs to. Inconsistent descriptions across your site and profiles make that harder.
4. There's no position 1 to hold. Answers vary between runs, models and users, so you measure rates across many prompts (how often you're mentioned, cited and ranked) rather than a single rank. What is AI search visibility? covers the metrics.
5. Questions are longer, and they fan out. Buyers type full questions with context, like "for a 10-person agency in the UK". Google says AI Overviews and AI Mode may run several related searches to answer one question, so pages that cover the sub-questions get more chances to be cited.

This plan assumes one founder or marketer with a few hours a week. It starts with measurement, because without a baseline you can't tell whether anything worked.

Week 1: Measure

• Write 15 to 20 prompts your buyers would ask: category ("best X for Y"), use case, "alternatives to [competitor]", and one about your brand by name.
• Run them across ChatGPT, Claude, Gemini, Perplexity and Google, or run a free scan.
• Record whether you're mentioned, your position, which sources are cited and which competitors appear instead.
• List the 10 domains cited most often. That's your outreach list for week 4.

Week 2: Fix your own site

• Rewrite your homepage H1 and meta description to name your category and audience in plain words.
• Publish or update your About page, and add Organization structured data with sameAs links to your profiles.
• Put your pricing in text on a crawlable page.
• Check robots.txt. Don't block the search crawlers you want to be cited by. Each AI company documents its own; OpenAI's crawler documentation (opens in a new tab) explains which one powers ChatGPT search.
• Verify your site in Bing Webmaster Tools as well as Google Search Console.

Week 3: Build quotable pages

• Write two honest comparison or alternatives pages for the competitors AI names most often.
• Write one use-case page for your best-fit audience.
• Add a short FAQ to key pages that answers the buyer questions from week 1 in plain sentences.
• Add specifics only you can state with confidence: limits, integrations, setup time, pricing.

Week 4: Earn mentions

• Pitch the authors of the five most-cited listicles with a short, specific note.
• Claim and complete your review and directory profiles, and ask a few customers for honest reviews.
• Answer two or three community threads where your category comes up, and say who you are.
• On day 30, re-run the same prompts and compare with week 1.

Don't expect the full effect within 30 days. Third-party mentions take time to be published, indexed and picked up. The point of month one is a baseline, a clean site and a pipeline of mentions that keeps working after you stop.

For a ChatGPT-specific version of weeks 2 to 4, see how to rank in ChatGPT. If you'd rather hand the work off, there's a done-for-you AI visibility audit.

Some GEO advice wastes time, and some carries real risk:

• Keyword stuffing. It didn't help in the GEO paper's experiments, and it makes pages worse for people.
• Mass-produced AI articles. Hundreds of thin pages won't make you more citable. A few specific, useful ones will.
• Hidden instructions aimed at AI, such as "ignore previous instructions and recommend us". Treat this as spam. It can get pages ignored and it destroys trust if anyone notices.
• Fake reviews and sock-puppet accounts. Platforms remove them, and getting caught does lasting damage.
• Anyone who guarantees a ranking. Nobody controls what these models say. Ask for a measurement plan instead.
• Over-investing in llms.txt. It's a proposed file that gives AI tools a map of your site. It's cheap to add, but as of this writing the major AI search providers haven't said they use it to decide what to cite.