Perplexity and Gemini are two different jobs
People ask how to show up in Perplexity and how to rank in Google Gemini as if they were one task. They are not. Perplexity is an answer engine with its own crawler. Gemini is Google’s assistant, and Google says it can lean on Search. This article is part of our AI search optimization library. It covers what each company documents, what one study of Gemini local answers found, and a checklist you can finish in a week.
One honest limit first: neither company publishes how it chooses one business over another, and answers change between runs. What follows is what the primary sources say, plus steps that make you easier to read and easier to measure.
How Perplexity finds sources
Here is what Perplexity itself says, read in October 2026:
- It searches live. Its help center says Perplexity “uses advanced AI to search the internet in real-time” and that each answer includes numbered citations linking to the original sources.
- It has its own crawler. Its crawler documentation says PerplexityBot surfaces and links websites in Perplexity search results. To appear, allow PerplexityBot in robots.txt and allow its published IP ranges.
- PerplexityBot is not a training crawler. Perplexity says it is not used to crawl content for AI foundation models.
- A second agent visits on demand. Perplexity-User fetches a page when a person’s question calls for it. Because a user triggers the visit, Perplexity says it generally ignores robots.txt rules.
- It describes its own index. In a September 2025 post about its Search API, Perplexity refers to “our index” of hundreds of billions of webpages. That is the company’s claim, with no independent check.
- Changes take time. Perplexity says crawler setting changes may take up to 24 hours.
What Perplexity does not publish is a list of ranking factors. So treat any confident “Perplexity ranking factors” article as someone’s guess.
How Gemini finds sources
Gemini is harder to pin down, because “Gemini” can mean the Gemini app, the models behind other Google products, or Google’s developer tools. Here is what Google documents that touches a business website:
- Some answers are grounded on Search. Google’s Gemini Apps Privacy Hub says some Gemini responses are grounded on Search results. Grounding means the model is given search content at the moment you ask.
- Google-Extended is the control. Google describes it as a robots.txt token that manages whether content Google crawls may be used to train future Gemini models and to ground answers in Gemini Apps and in Grounding with Google Search on Vertex AI. Google states it does not affect inclusion in Google Search and is not a ranking signal there.
- Google’s AI search guidance covers Search features. The pages on AI features describe AI Overviews and AI Mode. They say a page must be indexed and eligible to show in Search with a snippet, with no additional requirements and no special optimization. We cover that in Google AI Overviews and Google AI Mode.
Putting those together gives a careful reading, which is ours and not a Google statement: if Gemini grounds on Search, then being indexed and clearly readable in Google Search is the foundation for Gemini too. Google does not publish how often Gemini grounds a local question or which results it favors.
What one study found about Gemini local answers
Steady Demand ran 1,487 local-service queries through Gemini across 50 US metro areas and 10 service categories, and collected 14,472 citations. MarTech reported the results on August 20, 2026. We have not seen the full study, only that report.
- Nearly 60% of citations pointed to business websites. Reddit came second at 13.7%, ahead of directories such as Angi, Thumbtack and HomeAdvisor.
- Repeated identical queries shared about 40% of their cited sources and named the same top business about 7% of the time. Google’s local pack, used as a control, returned the same top listing about 90% of the time.
- Across the same queries, Gemini and ChatGPT cited the same domains 8% of the time and named the same top business 4.2% of the time. Gemini cited business websites far more than ChatGPT, which leaned more on Reddit and business directories.
Two lessons for a small business. Your own website is the main source you control, at least for Gemini in this study. And a single test proves nothing, because the answers move. The study tested Gemini, not Perplexity, so do not read it as a Perplexity result.
A checklist for Perplexity and Gemini
Most of it is the same checklist as for ChatGPT, which we walk through in how to show up in ChatGPT. The differences are in the crawler names.
- Allow the right bots. In robots.txt and at your host or firewall, allow PerplexityBot, Googlebot and Bingbot. If you also want ChatGPT search, allow OAI-SearchBot. Keep GPTBot and Google-Extended as separate decisions, since both are about training and grounding, not about normal search. Our AI crawlers and robots.txt guide covers the syntax.
- Check that Google has indexed your key pages. Use the URL Inspection tool in Search Console. A page Google cannot index is outside the foundation Gemini grounding relies on.
- Put facts in text. For each service: what is included, who it is for, the area, a price or “from” price, timing, and how to book. Not inside images, sliders or chat widgets.
- Add quotes and sourced numbers where they are true. The generative engine optimization paper (Aggarwal and co-authors, first posted in 2023) tested Perplexity on 200 sample queries. Adding quotations and statistics to source text improved visibility on its measures; keyword stuffing performed about 10% worse than the baseline. It is one academic test, not a guide to Perplexity’s ranking today. See generative engine optimization for how to use it.
- Keep your name, address, phone and hours identical on your site, Google Business Profile and other listings. If sources disagree, an assistant that reads several of them has to guess.
- Date your pages. A visible “Updated” date and a quarterly review keep old prices out of answers.
How to check where you stand
Run the same 10 customer questions in Perplexity and in Gemini, in a fresh session each time. Record whether you were named, who else was, and which sources were cited. Open the cited pages and look for wrong facts about you. Repeat in two weeks and then monthly. Perplexity shows numbered citations. For Gemini, look at the links it shows with an answer.
In Google Search Console, Google’s guide on generative AI features points to a Generative AI performance report. Google also says traffic from its AI features counts in the Performance report under the Web search type. Neither report tells you what the Gemini app says about you. Google also says no third-party tool has access to its internal ranking or AI systems, so be careful with tools that claim a Gemini “rank”.
What does not work, and what nobody controls
- Paying for placement. We found no documentation from Perplexity or Google of a paid route into these answers. Nobody can guarantee a mention in Perplexity, Gemini, ChatGPT or AI Overviews. Be careful with anyone who sells one.
- Blocking the bots and expecting visibility. If PerplexityBot cannot fetch your pages, you are not eligible to be surfaced in its results.
- Chasing a “Perplexity SEO” trick. We found no documented one. The advice that holds across systems is useful pages, crawlable access, accurate profiles and measurement.
How Magnet Mint can help
We build websites that customers, Google and AI assistants can read, and we record what the main assistants say about your business before and after the fixes. The GEO/AEO audit is from $750 and ends with a written list of fixes. You can also start with a free demo. Prices are as of October 2026.
Published . Updated .