Gemini and Perplexity SEO: How to Get Cited by Both

Here that means getting your pages cited in each engine’s answers, not using Gemini as an SEO tool. The short version: Gemini mostly rides on Google’s index, and Perplexity mostly rides on one crawler you can check in five minutes. The rest of this guide shows the documentation behind both, grades seven tactics by how well the evidence holds, and ends with two questions to put to any agency.

How Perplexity chooses its sources (Perplexity SEO)

Perplexity answers with numbered links to the pages it used, which anyone can see by asking it a question. What Perplexity actually documents is the crawler layer: PerplexityBot exists “to surface and link websites in search results on Perplexity”, according to Perplexity’s crawler documentation. So the one documented eligibility lever is access. If robots.txt, a CDN rule or a firewall blocks PerplexityBot, your pages drop out of the search layer those links are drawn from.

The same documentation says PerplexityBot is “not used to crawl content for AI foundation models”. So allowing it is a search decision, not a training decision. The minimum PerplexityBot robots.txt entry is two lines: User-agent: PerplexityBot followed by Allow: /. Perplexity also publishes the bot’s IP ranges, and notes that robots.txt changes can take up to 24 hours to take effect.

There is a second agent. Perplexity-User fetches a page when a user’s question triggers it, and Perplexity’s docs say it “generally ignores robots.txt rules” because a person asked for the fetch. Robots.txt cannot hide a page from that kind of request.

Perplexity does not publish how it ranks or chooses between eligible pages. Every “Perplexity ranking factors” list you will find online is a vendor’s inference. Some of those inferences may be right. None of them is documented.

How Gemini chooses its sources (Gemini SEO)

Gemini can ground its answers in Google Search, so ordinary Google indexability and ranking carry over into Gemini SEO. According to Google’s grounding documentation, the model decides whether to search, writes one or more queries, and returns text with url_citation annotations linking to sources. In practice, SEO for Gemini starts with being found and ranked by Google.

That grounding page documents the developer API. The consumer Gemini app is less documented, so we know the app can draw on Google Search, not exactly how it picks each link.

Google-Extended is the part people get wrong. Per Google’s common crawlers documentation, it is a robots.txt token with no separate user-agent string. It controls whether your content “may be used for training future generations of Gemini models” and for grounding in Gemini Apps and in Grounding with Google Search on Vertex AI. Google states it “does not impact a site’s inclusion in Google Search” and is not used as a ranking signal in Google Search.

Google’s AI Overviews and AI Mode sit next to Gemini but follow Search rules. Google’s AI features guidance says a page must be indexed and eligible to show with a snippet, with “no additional requirements” and no special optimisations needed. Those features use “query fan-out”, issuing multiple related searches across subtopics. That describes AI Overviews and AI Mode, not the Gemini app.

Gemini vs Perplexity SEO: what’s the same and what’s different

Gemini SEO and Perplexity SEO share one requirement: the engine must be able to reach your page. They differ in who controls that. For Gemini, reach comes through Google Search, and the Google-Extended token decides only whether Gemini may use your content, according to Google’s crawler documentation. For Perplexity, reach depends on PerplexityBot, which you allow or block in robots.txt, according to Perplexity’s crawler documentation.

We left out a freshness row. Vendor guides claim both engines favour recent content, but neither engine documents it, and the vendors disagree on how much.

The Gemini work is “be good at Google”. The Perplexity work is “check you have not blocked a bot”. Both fit in one afternoon, so do them before anything else. If the GEO and AEO labels are new to you, our explainer on GEO vs AEO vs LLM SEO covers them.

How to rank on Perplexity and Gemini: 7 tactics

To rank on Perplexity and Gemini, start with three steps the engines themselves document: allow PerplexityBot, stay indexable and snippet-eligible in Google, and make a deliberate Google-Extended choice. Four more tactics are inference or common vendor advice with no primary proof. We found no peer-reviewed study of how either engine selects sources (searches, 3 November 2026), so treat the second list as reasonable bets.

Documented by the engines

1. Don’t block PerplexityBot. It is the eligibility lever for Perplexity’s search results, per Perplexity’s crawler docs. Check robots.txt first. Then, as our own advice, check any CDN or firewall bot rules against the bot’s published IP ranges, because a security setting can block it silently while your robots.txt looks fine.
2. Be indexable and snippet-eligible in Google. Google’s AI features guidance makes this the entry ticket for AI Overviews and AI Mode, and Gemini grounding draws on Google Search. Your first diagnostic is Google Search Console: is the page indexed, and is anything (a nosnippet tag, for instance) stopping a snippet?
3. Decide on Google-Extended on purpose. Blocking it keeps your content out of Gemini training and grounding in Gemini Apps, and leaves your Search rankings alone. Make the call on whether you want Gemini to use your content, not out of fear for your rankings.

Commonly recommended, not proven

4. Cover the subtopics, not just the head term. Google documents query fan-out for AI Overviews and AI Mode. Covering subtopics is our inference from that, and fan-out is not documented for the Gemini app or Perplexity. A page on office cleaning contracts that also answers pricing, contract length and what is included gives those related searches more to find.
5. Answer first. Put the direct answer in the opening lines of each section, so a passage can be lifted whole. If a buyer asks how much a service costs in Manchester, the first sentence under that heading should state the price range.
6. Keep pages dated and your details consistent. Show a visible update date on every guide and service page, and use the same business name, address and service list on your site, your Google Business Profile and every directory listing. Organization schema on your homepage states those details in a form machines read directly.
7. Earn third-party mentions. Being named on other credible sites is the most repeated advice in GEO. For a B2B service firm, that means trade association listings, supplier case studies and industry press that name you alongside the service you sell.

Ignore the precise numbers in other guides: schema as a fixed share of ranking, a word-count sweet spot, backlink thresholds for cited pages. Vendor studies contradict one another on these points, which tells you how unsettled the evidence is. A guide that prints them as rules is selling certainty it does not have.

How to check whether you’re being cited

The simplest way to check whether Gemini or Perplexity cites you is to ask them. Write 10 to 15 prompts your real buyers would type, such as “best commercial cleaning contractors in Leeds for offices”, run each one in both engines, and record whether you are cited and who is cited instead. Repeat monthly, because answers vary from run to run. This is our practical advice, not a documented method.

A spreadsheet is enough. Use six columns: prompt, engine, date, cited (yes or no), who was cited, and the cited URL. If an engine shows no sources at all for a prompt, record that too. Keep the prompts fixed so the months compare, and note when you changed a page so you can see whether the change moved anything. Any agency selling GEO should be able to show you the same log for its own site.

Do GEO agencies optimise for Gemini and Perplexity?

Of 271 GEO agencies GEO Agency Directory hand-verified, 109 (40%) name Perplexity on their service page and 66 (24%) name Gemini. Only 56 (21%) name both, and 152 (56%) name neither, according to our master data re-counted on 3 November 2026. Every one of the 271 names ChatGPT. The figures record what a service page names, not what an agency can do.

By headquarters (same master data), the gap is widest in the UK. Only 8 of 49 UK agencies (16%) name Gemini on their service page, against 18 that name Perplexity. Of 149 US agencies, 52 name Perplexity and 31 name Gemini. Singapore is the one market where Gemini is named more often than Perplexity: 9 of 26 agencies against 7.

The hand-verification method is public. Each agency’s service page was read and the AI platforms it names were recorded. We did not test any agency’s results in either engine. Our visibility score measures Google rankings from a fixed query panel, not AI citations, and we hold no citation dataset.

A service page that names only ChatGPT does not prove a gap in capability. We cannot tell from a service page why an engine is missing. It is a reason to ask. Put two questions on the call: “Which engines do you test, and can you show me a Perplexity citation you won?” and “Can I see your citation log for your own site?” For what else a good scope should cover, see what AI SEO services include.

Gemini and Perplexity SEO questions

Do I need separate content for Gemini and Perplexity?

No. Neither engine documents a reason to write separate pages. The same page serves both, as long as it passes two access checks: indexed in Google, and reachable by PerplexityBot.

Is Perplexity SEO different from Google SEO?

Partly. Perplexity uses its own crawler, so a page Google ranks can still be missing from Perplexity if PerplexityBot is blocked. The fundamentals of crawlable pages with direct answers overlap, and anything beyond that is unproven.

Which engine should I work on first?

Do the two cheapest checks first: Google indexation in Search Console, and PerplexityBot access in robots.txt and your firewall. Both checks take minutes, though fixing a problem may not, and together they cover the documented ground for both engines.