Answer engine optimization

How to get cited by Gemini using Claude

Gemini does not recall your brand from memory. It retrieves a handful of pages and writes its answer from them, so getting cited means getting onto those pages, in a form it can lift. With Claude, the workflow is: find the buyer questions where Gemini names a competitor instead of you, look at which URLs it keeps citing for those questions, and publish a page that out-answers them - leading with the answer, structured as a table or list, and honest about where competitors genuinely win. Then re-measure a few weeks later, because publishing without re-measuring is guessing.

How Gemini actually decides who to name

Gemini grounds its answers in Google Search results. That makes it the engine where classic SEO still carries the most weight - if you are already surfacing in Google for the query, you are in the candidate set Gemini writes from. The grounding step then decides which of those candidates it actually names.

What moves the number on Gemini

1. 1Rank in Google for the question first. Gemini's grounding pulls from the same index, so traditional SEO is not dead here - it is the entry ticket to the room where the AI answer is written.
2. 2Structure the page so the answer is liftable: headers that match the question, a summary that stands alone, schema markup where it fits. Grounding rewards pages a machine can parse confidently.
3. 3Earn mentions on the sites Google already trusts for the topic. Grounding disproportionately names entities that appear across multiple retrieved sources, so being named on three pages beats owning one.

Doing it with Claude

An AI assistant that can search the web, read pages, and write long-form content. With the Fulcru skill installed, it runs the whole AEO loop conversationally.

1. 1Install the free `fulcru` Claude Skill. It teaches Claude the playbook: audit a domain, build the buyer question set, find who gets named instead of you, and write the page that closes the biggest gap.
2. 2Ask: "Audit example.com for AI search visibility." Claude fetches the site, works out what it actually sells and who its real competitors are, and writes 10-20 questions a buyer would ask an assistant *before they have heard of you* - which is the only kind of question worth testing.
3. 3Have Claude search each question and record who gets named and which URLs keep appearing. Those recurring URLs are the pages AI reads in your space, and they are the real target.
4. 4Pick the one gap with the highest buying intent that you could honestly win, and have Claude write the page: answer in the first 100 words, comparison table naming competitors honestly, specific steps with real numbers.
5. 5Publish it. Then re-run the same question set a few weeks later and compare. The comparison is the only thing that tells you it worked.

What this will not do

The part everyone skips

Re-measuring. A published page that you never check is a belief, not a result. Run the same question set again a few weeks later and compare: were you named in 2 of 10 answers before, and 5 of 10 after? That number is the only thing that tells you the work landed - and when it does not move, that is information too. It usually means the answer set is dominated by a source you have not gotten onto yet.