How to Rank in AI Search (Google AI Overviews, ChatGPT, Perplexity)

1. What AI search is and why it changed

Google launched AI Overviews broadly in 2024 and by 2026 they appear on a large share of informational queries — the kind of "how to", "what is", and "best way to" questions that used to send people to a list of blue links. At the same time, Perplexity has grown into a serious AI-native search engine, and ChatGPT's web search has made it a real alternative for research. The result is that for many questions, the top result is now a synthesised AI answer that cites sources — not a ranked list of pages to click through.

This doesn't mean traditional SEO is dead. It means the bar has changed. Google's own guidance on AI Overviews makes clear that the underlying ranking signals still apply — but the content that gets pulled into the answer box is selected by different criteria than what gets ranked in position one.

2. What AI search engines look for

To be cited in an AI answer, your content needs to do three things: directly answer the question being asked (not dance around it), be structured so the AI can find and extract that answer quickly, and come from a site with enough credibility to be trusted. The mental shift is from "how does this page rank for this keyword?" to "can an AI extract a clear, direct answer from this page?"

• Directness: AI systems reward pages that lead with the answer, not pages that build slowly to it after preamble.
• Structure: Clear headers, short paragraphs, and formatted lists make content easy to parse and quote.
• Authority: Being linked to from other credible pages signals that your answer can be trusted.
• Freshness: AI Overviews favour recent, up-to-date content — stale pages get passed over even if they once ranked well.

3. Content structure AI can parse

The single biggest change you can make is to answer the question in the first paragraph, not after several hundred words of context. AI systems scan for the direct answer and pull it; pages that bury the answer don't get cited even if the rest of the content is excellent. Beyond that:

• Use headers as questions or clear statements — "What is first-party data?" is more parseable than "Background."
• Keep paragraphs to 2–3 sentences. Dense walls of text don't get pulled into AI answers.
• Use structured lists for steps, options, and comparisons. AI systems reproduce lists well because they're already formatted.
• Include a summary or definition near the top of pages that explain a concept — that summary is often the line that gets cited verbatim.

4. How to appear in ChatGPT and Perplexity

Both ChatGPT (with web search) and Perplexity actively cite the pages they draw from — which means appearing in their answers is directly traceable in your analytics via referral traffic. The signals that matter:

• Be indexed. Pages that haven't been crawled can't be cited. Submit new content to Google Search Console and check indexing status.
• Earn backlinks to your best content. Perplexity in particular rewards pages that are cited elsewhere on the web — a small number of quality backlinks to a thorough, direct-answer page can get it appearing in AI results consistently.
• Write about specific questions, not broad topics. A page titled "What is a good email open rate for SaaS in 2026?" is more citable than a page titled "Email Marketing Guide."
• Use schema markup. FAQ and Article schema help AI systems understand what your content is about and what questions it answers.

Restructuring the marketing.keyforriches.com guide pages to lead with a direct answer rather than a long intro improved their appearance in AI Overviews noticeably within a few weeks — the pages that got cited were consistently the ones where the answer was in the first two sentences, not buried in section three.

5. Google AI Overviews: what works in 2026

Google draws AI Overviews from its existing index, so the fundamentals — crawlability, indexing, E-E-A-T signals — still apply. What's specifically working for AI Overview inclusion in 2026:

• Named authors with demonstrated expertise. Pages with a clear author byline and bio outperform anonymous pages — Google's Article schema documentation shows how to signal authorship properly.
• Original data, experience, or perspective. AI Overviews avoid pulling generic summaries when a page with first-hand insight is available.
• Content that matches the exact search intent. A transactional query and an informational query need different page structures — mixing intent on one page reduces the chance of being pulled into either kind of answer.
• Fast-loading, mobile-friendly pages. Core Web Vitals remain a factor in what gets indexed fully and quickly.

The good news is that most of what makes content excellent for AI search — clarity, directness, demonstrated experience — also makes it better for human readers. A Readability & Clarity check on your most important pages will surface exactly where the writing gets too vague or slow to be useful for either.

Clear content works better in AI search — and in every other format too. See how to make your marketing copy clearer for the editing moves that matter most, and how to write headlines that convert for getting the structure right from the top.