Show up in the answers, not just the links.

SEO optimizes for blue‑link rankings — getting your page to show up in Google's main results. GEO optimizes for citations in AI answers — getting cited when ChatGPT, Perplexity, or a Google AI Overview answers a question your buyer is asking. The work overlaps a lot: same schema, same technical baseline, same content quality, similar E‑E‑A‑T signals. The differences are in optimization target, measurement, and how content gets formatted for clean extraction by an LLM. Most clients end up running both layered.

Yes. AI engines pull from the same web SEO has always optimized. The pages that rank well in Google are usually the same pages that get cited in AI answers, because the underlying signals overlap heavily. The work that wins blue‑link rankings — quality content, clean technical foundation, authoritative backlinks, real expertise — is still doing most of the heavy lifting in GEO. The new work is structural: schema density, extractable formatting, AI bot access, brand mention building. Layered, not replaced.

Three things have to be true. ChatGPT has to be able to crawl your site — check OAI-SearchBot in robots.txt and any server-level bot blocks. Your content has to be cleanly extractable — clear H2 questions, definitive answer paragraphs, schema markup that tells an LLM what type of content this is. And your brand has to appear in third-party sources ChatGPT trusts — directories, trade press, review platforms, sometimes Reddit. ChatGPT cites businesses with presence on multiple authoritative surfaces.

A repeatable check where I ask the major AI engines (ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) a set of questions a real buyer in your category would ask, and log which businesses get cited. Run quarterly or monthly depending on the engagement. It tracks whether you're showing up at all, what you're getting cited for, and how citation share is moving relative to competitors. Not the only measurement. The closest thing to a SERP rank tracker that exists for AI search right now.

llms.txt is a proposed standard, similar to robots.txt, that tells AI engines which pages on your site to prioritize and what each page contains. It's not officially adopted by the major engines yet, but Anthropic has expressed support and several smaller engines parse it. I include an llms.txt audit and setup as part of GEO engagements when it fits. Small lift, potential meaningful upside if the standard takes hold. If it doesn't, no real downside.

Yes. The work isn't gatekept. If you have an in-house developer who can implement schema, a content team that can rewrite for structured extraction, and time to set up GA4 referral tracking and run prompt audits — you can run GEO internally. Where I help: when those resources don't exist, when you want a specialist who's done this across multiple sites, or when you want a layered SEO + GEO program that doesn't fragment ownership across teams.

Schema and AI bot access changes can show up in AI citations within two to four weeks of crawl. Brand mention building runs longer — three to six months minimum. Structured-answer formatting compounds over time as you rewrite or publish more content in extractable shape. AI search measurement is also noisier than Google Search Console; I'd budget 60 to 90 days before claiming results are real and not crawl variance.

Helpful, not required. AI engines pull from a wider source set than Google's top 10 — they cite directory listings, Reddit threads, trade press, and other surfaces that may not rank in traditional SEO. That said, sites that rank well in Google tend to also get cited in AI search, because the underlying signals overlap. If your SEO foundation is weak, the audit usually surfaces SEO + GEO opportunities together — and we sequence them so the foundational SEO work is doing double duty.