Gemini SEO: how to get cited in Gemini answers

Gemini SEO is the work of getting your pages cited, and your brand named, when someone asks the Gemini app a question. Gemini grounds its answers in Google's Search index, so a page Google has not indexed is unlikely to be cited. To show up in Gemini, get the page indexed in Google, leave the Google-Extended token unblocked in robots.txt, and put your key facts in short passages that answer one question and name your brand.

Google has not published how Gemini ranks the sources it cites, and its guide to generative AI search covers AI Overviews and AI Mode without mentioning the Gemini app. So work from what Google has documented about grounding and Google-Extended.

How does Gemini find and cite websites?

Gemini pulls content from the Google Search index when a question arrives and writes its answer from that content. Google calls this grounding. Its common crawlers documentation lists Gemini Apps among the products that ground answers this way, to make them more factual and relevant.

Google has not documented the app's retrieval steps. The nearest description is its developer guide to Grounding with Google Search in the Gemini API. There the model decides whether a search would improve the answer. If it would, the model writes one or more queries, runs them and links each segment of its answer to the source URL behind it.

If the app works the same way, three things follow for a site owner.

• A question answered without a search cites nobody. Brand mentions in those answers come from training data.
• Your page competes for the model's own queries, which may not match the user's wording. Our guide to query fan-out covers how to answer those sub-questions.
• Citations attach to segments, so the unit that wins is a sentence or short passage that carries the fact on its own.

In the app, the Sources button sits at the bottom of a response or inline and opens the links in a side panel. Google's Gemini Apps help says not every response includes sources.

Gemini search vs AI Mode

Google Gemini search, meaning answers in the Gemini app, is a separate product from AI Mode and AI Overviews, with separate controls, even though all three read Google's index. AI Mode and AI Overviews run inside Google Search on a custom version of Gemini, which is why people mix them up.

The controls do not cross over. The Search generative AI setting covers AI Overviews, AI Mode and Discover's generative features, and its help page does not mention the Gemini app. Google-Extended has no effect on Search. A citation in AI Overviews tells you little about Gemini, and the reverse holds too. For the Search side, read AI Overviews vs AI Mode.

What Google-Extended does and does not do

Google-Extended is a robots.txt token. It decides whether Google may use content it crawls from your site to train future Gemini models and to ground answers in Gemini Apps and in Grounding with Google Search on Vertex AI. Google's crawler documentation says it does not affect inclusion in Google Search and is not a ranking signal there.

It is not a crawler. It has no user agent string of its own, and Google's existing crawlers do the fetching, so you will never see it in your logs.

One token covers training and grounding together. This group opts you out of both:

User-agent: Google-Extended
Disallow: /

With it in place, AI Overviews and AI Mode can still cite you, because Googlebot rules govern them. The Gemini app stops grounding answers in your content, and so does Grounding with Google Search on Vertex AI, which businesses use to build on Gemini.

If you want buyers to hear about you in Gemini, leave Google-Extended unblocked. I would block it only where the content is the product, as on a news site, and only as a deliberate choice. If you once pasted a "block AI bots" list into robots.txt, look for this group before you conclude that Gemini ignores you. Our robots.txt checker for AI crawlers shows whether your file blocks Google-Extended and the exact line to change.

Gemini SEO starts with Google indexing

Google documents no Gemini index and no way to submit pages to Gemini, so a page must be in Google's index before Gemini can ground an answer in it. Open the page in Search Console's URL Inspection tool and confirm the URL is on Google. If it isn't, look for a noindex tag, a canonical that points elsewhere, a robots.txt rule that blocks Googlebot, or a thin page Google chose to skip.

Google has not said whether Gemini favors pages that rank higher in Search. Treat indexing as the gate you can check and ranking as an open question.

How to rank in Gemini

Gemini's ranking signals are not public. The changes below help a page get picked for one of Gemini's searches and get the citation tied to your brand.

Answer one question per passage

Grounded Gemini answers link segments of text to sources, at least in the API, so write sections whose first sentence carries the whole fact with your brand name in it. Put the question in the heading, in the words buyers use.

<h2>Does Ledgerly work with Stripe?</h2>
<p>Yes. Ledgerly imports Stripe payouts, fees and refunds every hour
and matches each payout to its invoice. Stripe sync is on every plan.</p>

Write "Ledgerly", not "it", in that first sentence, so a lifted segment still says whose product it is. Give price, limits and integrations their own sections too, so each sub-question has a passage to land on.

Keep key facts in the server HTML

Among AI engines, Gemini is the exception on JavaScript. Vercel's 2024 study of AI crawlers found that none of the major AI crawlers rendered it, except Gemini, which uses Googlebot's infrastructure.

Server HTML still wins. Google's JavaScript SEO guide says a page can wait in the render queue for a few seconds or longer, and it recommends server-side rendering or pre-rendering. A price in the HTML does not wait for that queue, and it also reaches ChatGPT and Perplexity, which do not render. Don't rebuild a site for Gemini alone, but keep the facts you want cited out of client-side scripts.

Make the brand unambiguous in Google

When Gemini grounds an answer about your company, it works from what Google's index holds about it. Search your brand name on Google. If page one mixes you up with a namesake or shows an old name, a grounded answer about you is likely to start from the same confusion.

Use one name on your site, your social profiles and the directories that list you, and add Organization markup to the home page. Google says structured data isn't required for its generative AI features and has not said Gemini reads it. The markup earns its place because Google's Organization documentation says it can help Google disambiguate your organization in Search, and Search is the index Gemini grounds in.

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Ledgerly",
  "url": "https://ledgerly.example",
  "logo": "https://ledgerly.example/logo.png",
  "sameAs": [
    "https://www.linkedin.com/company/ledgerly",
    "https://github.com/ledgerly"
  ]
}

Our Brand Entity Checker tests whether your name matches across Organization and WebSite markup, og:site_name and the manifest, whether your sameAs profiles exist, and whether a Wikidata item lists your site.

How to check your Gemini citations

You have to ask Gemini, because Google gives site owners no data on the Gemini app. Search Console's Generative AI performance report counts impressions in AI Overviews and AI Mode only. Server logs can't show a grounded answer either, because grounding draws on content Google already crawled.

1. Write down 10 to 15 questions a buyer asks before choosing a product like yours.
2. Ask each one in a new Gemini chat and open Sources. Note whether your domain appears and which domains appear instead.
3. Flag questions that never get a Sources button. Those answers carry no links, so citation work has nothing to win there.
4. Repeat every few weeks and act on the domains that keep coming back, because answers change between runs.

The domains Gemini cites in your place are your next lead. If a comparison article it keeps citing names you, that can get you into the answer before your own page is picked.

Clicks from the Gemini web app usually arrive with gemini.google.com as the referrer. In GA4, open Reports, then Acquisition, then Traffic acquisition, set the dimension to Session source and search for "gemini". Visits with no referrer land in Direct, so treat the count as a floor. Our AI referral tracking guide has a regex that groups Gemini with ChatGPT, Perplexity and the other assistants.

Check whether Gemini names and cites your brand

Enter your site and one question a buyer would ask. AI Answer Visibility asks Gemini, ChatGPT and Perplexity that question with web search on. For each engine it reports whether the answer names your brand and where it ranks among the brands named, and whether the answer cites your site. It also lists the brands and sites each answer recommends instead, and it says when an engine cited no source at all.

If you leave the brand name blank, the tool reads one from your home page. It does not cover Google AI Overviews, AI Mode or Claude. A run costs 120 credits. Each answer is a sample, so run your most important questions more than once before you act on a single result.