AI SEO: How to Show Up in ChatGPT, Perplexity, and AI Search

Start with accurate, accessible pages that answer your customers' questions. Support factual claims with sources and keep your business name, location, and services consistent. AI SEO is the practice of improving how a business can be found and represented in AI search. This guide covers practical checks used in MintUp's AI search optimization service, without promising placement.

This is not the same as using AI to do traditional SEO (though that is useful too). AI SEO, sometimes called Generative Engine Optimization or GEO, is about optimizing FOR AI search engines. Different game, different rules. Let's break down what works. This is also the exact discipline we run inside MintUp's AI search optimization service.

The GEO study by Aggarwal et al., presented at KDD 2024, reported visibility gains of up to 40% in its benchmark. It tested approaches including citations, statistics, and quotations. Results differed across domains, so the study supports testing specific editorial improvements rather than promising a fixed lift for a business.

Pew Research Center's analysis of 900 US adults found that in March 2025, users clicked traditional results on 8% of visits with an AI summary, compared with 15% without one. Links inside summaries were clicked on 1% of visits with a summary. This observational comparison does not establish that rankings stayed constant or that summaries caused the difference. Track citations and referral visits separately.

To measure your own site, record when substantive changes ship, then compare referral sessions and conversions over time. A visit from chatgpt.com shows a referral from that source. It does not establish that an llms.txt file or a rewritten paragraph caused the visit.

How is AI search different from Google search?

Google shows you 10 blue links and lets you pick. AI search engines like ChatGPT, Perplexity, and Google's AI Overviews read dozens of sources, synthesize the information, and give you a single answer with citations. The user never clicks through to most of those sources. They read the AI's summary, maybe click one or two citations, and move on.

This changes the game fundamentally. In traditional SEO, ranking on page one gets you clicks. In AI search, getting cited in the answer is the new page one. And the way AI decides which sources to cite is different from how Google ranks pages.

Which AI search platforms should you optimize for?

Optimize for the four that drive business queries: ChatGPT, Perplexity, Google AI Overviews, and Claude. Each runs its own crawler and cites sources in its own way, so the first practical step is confirming none of them are blocked in your robots.txt. The table below maps each engine to its crawler and the single move that does the most to win a citation there.

ChatGPT with Web Browsing

ChatGPT can search the web and cite sources in its answers. OpenAI's publisher guidance distinguishes OAI-SearchBot from its training crawler. Check robots.txt and hosting rules for search access; allowing access makes a page available for retrieval, without guaranteeing a citation.

Perplexity

Perplexity is a search engine built entirely on AI answers. Every query gets a synthesized response with numbered citations. Perplexity users are often researchers, decision-makers, and professionals comparing options. Its crawler (PerplexityBot) is separate from Google's. Perplexity tends to cite sources that are well-structured, factually specific, and recently updated.

Google AI Overviews

Google's AI Overviews appear at the top of search results for an increasing number of queries. They pull from Google's existing index but favor content that directly answers the query in a clear, concise passage. If you already rank on page one of Google, you have a shot at being cited in AI Overviews. But ranking alone is not enough. The content format matters.

Is on-page optimization enough to win AI citations?

No. Google's guidance for AI features says a page must be indexed and eligible for a snippet to appear as a supporting link. Normal SEO guidance applies, with no additional technical requirement or special AI markup. Meeting those conditions does not guarantee that Google will show the page.

Keep your own pages accurate and check how other sites describe your business. Reviews, directories, and publications can supply additional context. Their relevance varies by query and platform, so measure mentions instead of assuming a fixed formula for citations.

7 Practical Steps to Show Up in AI Search

1. Write Passage-Level Answers

AI search engines extract specific passages, not entire pages. When someone asks "what is AI SEO," the AI is looking for a 2-3 sentence passage that directly answers that question. Write clear, self-contained passages at the top of your content that answer the target query head-on. Front-load the definition or answer, then expand with details below.

2. Add Structured Data (Schema Markup)

Use structured data that matches the page: Article for articles, and a consistent Organization identity for the business. Check that the markup agrees with visible content. Google does not require a special schema type for its AI features.

Google retired FAQ rich results starting May 7, 2026. Visible questions and answers can still help readers. Keep existing markup accurate, but do not add it expecting a FAQ rich result or a guaranteed AI citation.

3. Create an llms.txt File

llms.txt is a proposal for a Markdown guide to a site's useful content. It can provide a concise overview for tools that choose to read it. It is optional, and adoption by a search engine should not be assumed. Google's AI-feature guidance says no new AI text files are needed for inclusion.

4. Build Entity Signals

AI search engines need to trust your brand as a legitimate entity. This means consistent information across your website, Google Business Profile, LinkedIn, industry directories, and anywhere else your brand appears. Your company name, location, services, and expertise should match everywhere. If your site says you are a "marketing agency" but LinkedIn says "creative studio" and Google Business Profile says "advertising agency," AI gets confused about what you actually are.

5. Demonstrate Topical Authority

AI engines favor sources that have depth in a topic area. One article about AI is not enough. A cluster of 5-10 articles covering different angles of AI for business signals that you actually know what you are talking about. Internal linking between related articles strengthens this signal further. This is content strategy, not content quantity.

6. Include Specific Numbers and Data

AI search engines love citing sources that include specific data points. "Our clients save time with automation" is invisible to AI. "Our clients save an average of 20 hours per week with CRM automation" is citable. Every claim you make should be as specific as possible. Percentages, dollar amounts, timeframes, and measurable outcomes make your content citation-worthy.

7. Keep Content Fresh

Keep dates honest. Correct outdated advice and update examples when the underlying facts change. A fresh date alone does not make a page useful or guarantee selection. Review time-sensitive articles regularly, and change the displayed update date when you make a substantive revision.

What can you do this week?

You do not need to overhaul your entire site. Here is a 5-step action plan you can execute in a few hours. If you want the deeper tactical breakdown of every signal AI systems weigh (entity building, review patterns, platform-specific strategies), read how to get AI to recommend your business next.

1. Check your robots.txt. Make sure you are not blocking OAI-SearchBot, PerplexityBot, or other AI crawlers. If your robots.txt blocks all bots except Googlebot, AI search cannot find you.
2. Audit your top 5 pages. For each one, add a clear 2-3 sentence answer to the page's primary question in the first paragraph. Make it quotable.
3. Put a sourced number behind your vaguest claim. Find the page that says you "save clients time" and replace it with the actual figure. This is the single most evidence-backed move on the list.
4. Optionally maintain an llms.txt guide to your company and key pages. Treat it as a useful index for compatible tools, not a proven search-ranking tactic.
5. Search for your brand in ChatGPT and Perplexity. Ask questions your customers would ask. See if you show up. If you do not, that is your baseline to improve from.

Frequently Asked Questions

Nick Vadini

CTO at MintUp

Nick is the full-stack engineer who architects and ships MintUp's builds out of Brunswick, Ohio, from infrastructure to frontend polish across React, React Native, Supabase, Stripe, and AI integrations. He has spent years building the AI systems, custom software, and automations that let Northeast Ohio businesses run leaner.

More about Nick Vadini