To show up in ChatGPT, Perplexity and Google's AI answers, your pages have to be reachable by each platform's search crawler, indexed, and then be the clearest, most specific source on the question being asked. That's less mysterious than the "GEO" industry makes it sound. Google says plainly that there are no special optimizations needed for its AI features, and none of the AI search vendors publishes a list of ranking factors.
So this guide sticks to what the platforms document themselves, and labels everything else as what it is: reasonable practice, not a secret formula.
What is GEO, and is it different from SEO?
Generative engine optimization (GEO) is the practice of making your content easy for AI search tools to find, understand and cite. In practice it's mostly SEO with a different scoreboard: instead of a blue link in position three, the goal is a citation or a mention inside an AI-written answer.
The overlap is big because these tools search the web before they answer. Google describes a technique it calls "query fan-out": AI Overviews and AI Mode may run several related searches across subtopics and data sources, then build the response from what they find. If your page doesn't rank or isn't indexed for any of those searches, it's not in the pool the answer is written from.
For a small business, this work usually sits inside a broader content and campaign plan. That's how we approach it in our AI marketing automation work: the same pages have to rank, get cited and convert.
What Google says about AI Overviews and AI Mode
Google's Search Central documentation, AI features and your website, is the most explicit statement any vendor has made, and it's refreshingly direct:
- To appear as a supporting link, a page must be indexed and eligible to be shown in Search with a snippet. There are no additional technical requirements.
- Google says there are no additional requirements or special optimizations needed to appear in AI Overviews or AI Mode, and you don't need new machine-readable files, AI text files or special schema markup for them.
- The recommended practices are standard SEO: allow crawling, make content findable through internal links, give visitors a good page experience, keep important content in text, support it with good images and video, make structured data match what's visible on the page, and keep Merchant Center and Business Profile information current.
- Traffic from AI features is counted in the Search Console Performance report, under the "Web" search type.
- To limit what's shown, you use the same preview controls as regular Search:
nosnippet,data-nosnippet,max-snippetornoindex. Google-Extended is a separate control for training and grounding in other Google systems, and it doesn't affect Search.
In May 2025, John Mueller of Google Search Relations summed up the content side on the Search Central blog: focus on unique, non-commodity content that people will find useful, because longer, follow-up-style questions in AI search reward depth.
There's one newer control worth knowing. On June 3, 2026, Google announced a Search Console setting that lets site owners opt out of its generative AI Search features (AI Overviews, AI Mode and AI Overviews in Discover), along with new reports on how pages appear in them. Google's post, updated August 31, 2026, says these features had rolled out to all websites worldwide. Google is clear that opted-out sites won't get traffic or impressions from those features, and that the setting isn't used as a ranking signal for regular results.
What OpenAI and Perplexity document
Neither OpenAI nor Perplexity publishes ranking guidance. What they do publish is how their crawlers work, and that's where most sites have problems.
Search crawlers vs training crawlers
| Criterion | ChatGPT (OpenAI) | Perplexity | |
|---|---|---|---|
| Crawler that decides search visibility | Googlebot | OAI-SearchBot | PerplexityBot |
| Training opt-out | Google-Extended (doesn't affect Search) | GPTBot | PerplexityBot isn't used for foundation models, per Perplexity |
| User-triggered fetcher | Not addressed in Google's AI features guide | ChatGPT-User: robots.txt may not apply | Perplexity-User: generally ignores robots.txt |
| Changes take effect | After recrawling, from days to months | About 24 hours, per OpenAI | Up to 24 hours, per Perplexity |
| Where to measure | Search Console, Web search type | Referrals tagged utm_source=chatgpt.com | Referral traffic in your analytics |
- Crawler that decides search visibility
- Googlebot
- Training opt-out
- Google-Extended (doesn't affect Search)
- User-triggered fetcher
- Not addressed in Google's AI features guide
- Changes take effect
- After recrawling, from days to months
- Where to measure
- Search Console, Web search type
ChatGPT (OpenAI)
- Crawler that decides search visibility
- OAI-SearchBot
- Training opt-out
- GPTBot
- User-triggered fetcher
- ChatGPT-User: robots.txt may not apply
- Changes take effect
- About 24 hours, per OpenAI
- Where to measure
- Referrals tagged utm_source=chatgpt.com
Perplexity
- Crawler that decides search visibility
- PerplexityBot
- Training opt-out
- PerplexityBot isn't used for foundation models, per Perplexity
- User-triggered fetcher
- Perplexity-User: generally ignores robots.txt
- Changes take effect
- Up to 24 hours, per Perplexity
- Where to measure
- Referral traffic in your analytics
The OpenAI details come from its Overview of OpenAI Crawlers: OAI-SearchBot is the crawler for search, and sites that opt out of it won't be shown in ChatGPT search answers, though they can still appear as plain links when someone looks for them by name. GPTBot is the training crawler, and blocking it signals that your content shouldn't be used to train foundation models. The two settings are independent. ChatGPT-User handles visits that a person triggers inside ChatGPT, isn't used for automatic crawling, and doesn't decide search visibility. OpenAI's publisher FAQ adds that ChatGPT search referrals carry utm_source=chatgpt.com, so you can see them in your analytics.
Perplexity's crawler documentation follows the same split: PerplexityBot surfaces and links websites in Perplexity's search results and isn't used to crawl content for AI foundation models, while Perplexity-User fetches pages when a user asks a question and generally ignores robots.txt.
What tends to earn citations (our practice, not vendor rules)
Here's the honest part: nobody outside these companies knows exactly how sources are chosen for a given answer. What follows is what we do because it matches what the vendors do say (unique, useful, well-structured, accessible content) and because it's good for human readers anyway.
- Answer first. Put the direct answer in the first two sentences of the page and under each main heading. AI tools quote passages, and a clear passage is easier to quote than a buried one.
- Be specific and first-hand. Real processes, real constraints, real examples from your work. Generic summaries of what everyone else says are exactly what Google calls commodity content.
- One page per question. A focused page on "how much does a water heater replacement cost in Austin" is a better source for that question than a general services page.
- Consistent facts about your business. Name, location, services and hours should match across your site, Business Profile and directories. Conflicting facts give an AI tool no reason to trust any of them.
- Name your sources. When you cite a statistic or a rule, name the organization and date. It helps readers, and it makes your page a better source.
- Text over images. Prices, service areas and policies should be in HTML text, not only in a PDF or an image.
You'll see vendors selling llms.txt files and special "AI schema." Google's documentation says neither is needed for its AI features. Other platforms haven't said they use them. They're cheap to add, but don't expect them to do anything on their own.
Technical GEO check
How do you measure AI search visibility?
Combine three signals: Search Console, referral traffic and regular manual checks. None is complete on its own.
- Search Console. AI feature traffic is part of the Performance report for the Web search type, and the generative AI reports Google added in 2026 show how your pages appear in those features.
- Analytics referrals. Track sessions from chatgpt.com, perplexity.ai and other AI tools, and look at what those visitors do: calls, forms, bookings. Google itself recommends judging visits by outcomes rather than clicks alone.
- Manual prompt checks. Keep a list of 20 to 30 questions your customers really ask and run them monthly in each tool. Answers change between runs, so look at trends over months rather than a single screenshot.
If you're a local business, Google's AI Mode and AI Overviews add a few specifics of their own, from Business Profile data to Google calling businesses on a user's behalf. We cover those in Google AI Overviews and AI Mode for local businesses.
When it makes sense to bring in help
Most of this is fundamentals done carefully. Where outside help pays off is in auditing what crawlers can reach, deciding which pages each question deserves, producing the first-hand content consistently and building the reporting. If you also want your own content to be searchable inside an assistant you control, that's a different project, closer to our LLM and RAG development work, and our explainer on what RAG is shows how AI tools retrieve passages before answering.
Appyweb has been building software since 2012, and we treat AI search as part of the same marketing system as your site, ads and CRM. If you'd like a review of how your business shows up in AI answers today, start with our AI marketing automation page or ask us for a visibility check.