Oliver Zeman
Does SEO Still Work With AI Search Tools Like ChatGPT?
SEO works with AI search tools like ChatGPT — but the rules changed. Here's what matters now.
You've heard the take. SEO is dead. ChatGPT killed it. Nobody clicks links anymore.
Here's what nobody tells you: every AI search tool you're worried about runs on the same infrastructure SEO built. ChatGPT pulls from Bing's index. Google's AI Overviews draw from the same top-ranking pages you're already competing for. Perplexity scans the live web. These tools don't create knowledge — they extract it from pages that search engines already trust.
The question isn't whether SEO works. It's whether your SEO works for the way search actually happens now.
Gartner predicted traditional search volume would drop 25% by 2026. They were right. But that same year, 88.5% of Australian websites started receiving AI-referred traffic — up 1,200% year over year. ChatGPT alone drives roughly 90% of that traffic. The pie didn't shrink. It split into more slices.
If your SEO strategy still treats Google as the only game in town, you're leaving traffic on the table. Here's what changed, what didn't, and how to show up where your customers are actually looking.
How AI Search Tools Actually Find Your Content
AI search engines use a technique called Retrieval-Augmented Generation — RAG for short. When someone asks ChatGPT "what's the best CRM for a small sales team," the model doesn't pull from its training data. It searches the live web, grabs relevant pages, synthesizes them, and cites sources.
The key word is retrieval. Before the AI can answer, it has to find something worth citing. That retrieval step is pure SEO.
Google's own documentation confirms this. Its AI features are "rooted in our core Search ranking and quality systems." The same crawlability, relevance, and authority signals that determine your Google ranking also determine whether an AI tool picks up your content.
But here's the twist: AI tools don't simply mirror Google rankings. Research from 3Aspens Media found that 90% of ChatGPT citations come from websites outside the top 20 traditional Google results. AI evaluates content relevance independently — it cares more about whether you answer the question than whether you bought the right backlinks.
This is the single biggest opportunity in search right now. You can rank on page three of Google and still get cited by ChatGPT if your content is structured well and directly answers real questions.
What each AI search engine actually uses
| AI Tool | Search Backend | Citation Behavior |
|---|---|---|
| ChatGPT | Bing index | Cites sources inline; favors direct answers |
| Google AI Overviews | Google index | Pulls from top-ranking pages; link cards below |
| Gemini | Google index | Often cites 10+ sources; synthesizes broadly |
| Perplexity | Live web crawl | Heavy citation; favors academic/authoritative sources |
| Claude | Web search mode | Fewer citations; prioritizes synthesis |
Each engine has its own preferences. But they all share one requirement: your content has to be findable, crawlable, and clearly structured.
What Still Works (And Always Will)
Some things haven't budged. The fundamentals that drove rankings a decade ago still determine whether AI tools find you today.
Technical SEO. If your site blocks crawlers in robots.txt, loads slowly, or buries content behind JavaScript that never renders, AI tools can't read you. Period. Google's new AI optimization guide explicitly recommends following crawling best practices, using semantic HTML, and maintaining good page experience. The same technical health that keeps you in Google's index keeps you in ChatGPT's retrieval pool.
Backlinks and authority. AI tools evaluate trust the same way search engines do — by looking at who else references you. A site with strong domain authority and a healthy backlink profile is more likely to be retrieved and cited. Google's documentation puts it plainly: core ranking systems focus on quality, and those systems feed AI features.
Clear, helpful content. AI models don't reward keyword stuffing. They reward content that answers a question completely and clearly. Google's guide draws a sharp line between "commodity content" — think "7 Tips for First-Time Homebuyers" — and non-commodity content that brings unique insight, like "Why We Waived the Inspection and Saved Money: A Look Inside the Sewer Line." The latter wins every time.
Local SEO. Google Business Profiles, reviews, and location signals still dominate local search. AI Overviews appear far less often in high-intent local queries like "plumber near me" or "emergency electrician Denver." If you run a local business, your Google Maps presence remains your most valuable digital asset.
The fundamentals aren't optional. They're the entry ticket. Without them, AI tools never see your content in the first place.
What's Different Now
The fundamentals haven't changed, but the shape of good content has. AI tools extract answers at the passage level, not the page level. That changes everything about how you write.
Direct answers win. An answer buried in paragraph twelve is invisible to AI. The same answer in the first two sentences under a clear heading gets cited. This isn't a minor formatting preference — it's the difference between being the source and being invisible. Brisbane SEO Expert found that content structured with the answer first, under a descriptive heading, gets cited at dramatically higher rates.
Questions beat keywords. AI search is conversational. People type "how much does it cost to fix a leaky roof in Phoenix" not "roof repair pricing." Content built around natural-language questions — the exact phrases your customers use — outperforms content optimized for short keywords. FAQ sections built around real customer questions carry more weight than they ever did in traditional SEO.
Entity authority matters beyond your site. AI tools don't just crawl your website. They check third-party mentions — Reddit threads, industry review sites, Wikipedia, news outlets — to evaluate whether you're actually trustworthy. A brand mentioned consistently across G2, Capterra, Reddit, and industry publications looks more credible to an AI than a brand with a perfectly optimized website and zero external footprint.
Zero-click is the norm. Over 80% of Google searches now end without a click. When AI Overviews appear, organic click-through rates drop to 0.52% for uncited sites — a 65% year-over-year decline. Sites that are cited in the AI Overview fare better at 0.70%, but that's still down 49%. The traffic you do get is more qualified, but there's less of it. Visibility now means being the source an AI cites, not just the link someone clicks.
Traditional SEO vs. AI search optimization
| Traditional SEO | AI Search Optimization | |
|---|---|---|
| Goal | Rank #1 for a keyword | Get cited in the AI's answer |
| User action | Click through to your site | Read the synthesized answer; click if curious |
| Content shape | Long-form, keyword-dispersed | Direct answers, clear headings, FAQ structure |
| Authority signal | Backlinks and domain rating | Backlinks + third-party brand mentions |
| Success metric | Rankings and organic clicks | Citations, AI-referred traffic, brand mentions |
The AEO/GEO Debate: Why Google Says It's All Still SEO
You've probably heard the acronyms. AEO — Answer Engine Optimization. GEO — Generative Engine Optimization. Entire agencies have rebranded around these terms. Consultants charge premiums for "GEO audits."
Google's position is unambiguous. In May 2026, the company published its official AI optimization guide with a section titled "AEO and GEO are still SEO." The documentation states: "From Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO."
Google also named specific tactics you don't need: llms.txt files, content chunking, AI-specific rewriting, and special schema markup. Several of these appear prominently in GEO guides and paid courses. Google's engineers explicitly recommend against chunking content into small pieces for AI systems — their systems "are able to understand the nuance of multiple topics on a page and show the relevant piece to users."
This doesn't mean GEO is a scam. It means the label is mostly marketing. The underlying practices — clear structure, direct answers, entity authority — are just good SEO applied to a new surface. The people who win aren't the ones who learn a new acronym. They're the ones who do the fundamentals better than everyone else.
See how Lookelo tracks your brand's AI citations across ChatGPT, Gemini, and Perplexity.
How to Structure Content So AI Tools Cite You
Five things move the needle. None of them require new tools or separate frameworks.
1. Put the answer in the first two sentences
Every section of your content should open with a direct, self-contained answer to the question the heading promises. AI tools extract these opening sentences verbatim. If your first paragraph is throat-clearing, the AI moves on to someone else's page.
Bad: "When it comes to choosing a CRM for your small business, there are many factors to consider in today's competitive landscape."
Good: "The best CRM for a small sales team under 10 people is HubSpot's free tier, because it includes email tracking, deal pipelines, and meeting scheduling at no cost."
2. Use descriptive, question-based headings
AI tools scan heading hierarchies to understand what a page covers. Vague headings like "Our Approach" or "More Information" tell the AI nothing. Specific headings like "How Much Does Local SEO Cost in 2026" or "What Is the Difference Between SEO and GEO" tell the AI exactly what the section answers.
This also makes your content more scannable for human readers — which is the point. AI optimization and user experience converge here.
3. Build FAQ sections that answer real questions
FAQ sections are the single most efficient way to get cited by AI tools. They're structured exactly how AI retrieval works: question → direct answer. Use the "People Also Ask" box in Google, Reddit threads, and customer support tickets to find the questions your audience actually asks. Answer each one in 2-4 sentences under an H3 that matches the question phrasing.
4. Use schema markup
Google says no special schema is required for AI features. But standard schema — FAQ, HowTo, Article, Product, LocalBusiness — helps all search engines understand your content. Schema is not a ranking factor for AI citations, but it removes ambiguity. When a crawler hits your page, schema tells it immediately "this is a product page with pricing, reviews, and availability" rather than making the crawler figure it out from context.
5. Keep your site open to AI crawlers
Check your robots.txt. OpenAI's crawler (GPTBot), Google's AI crawlers, and Anthropic's Claude crawler all respect robots.txt directives. If you're blocking them, you're opting out of AI visibility entirely. Most sites block AI crawlers by default through security plugins or overly aggressive bot filtering. A quick robots.txt audit can be worth thousands in recovered traffic.
The Metrics That Matter in 2026
Rankings aren't enough anymore. You need to track whether AI tools mention your brand, which pages they cite, and how much traffic those citations drive.
Citation monitoring. Tools like Lookelo track brand mentions across ChatGPT, Gemini, Perplexity, and Google AI Overviews. You can see which queries trigger citations, which pages get referenced, and how your visibility compares to competitors. This is the new rank tracking.
AI-referred traffic. Google Search Console and analytics platforms now segment AI-referred traffic. Look for referral sources from google.com/ai, chatgpt.com, perplexity.ai, and gemini.google.com. Traffic from these sources is typically lower volume but higher intent — someone who clicks through from an AI citation is actively researching, not casually browsing.
Brand mention velocity. How often does your brand appear on Reddit, Quora, G2, industry blogs, and news sites? AI tools use these third-party signals to evaluate trust. A steady increase in brand mentions across reputable platforms correlates with higher AI citation rates. This is word-of-mouth marketing with a measurable SEO payoff.
CTR by query type. Segment your organic CTR by whether AI Overviews appear for your target queries. If AI Overviews are present and you're not cited, you're losing 65% of potential clicks. Prioritize optimizing the pages targeting those queries for direct answers and clear structure.
A Practical 5-Step AI SEO Checklist
Audit your technical foundation. Crawl your site. Fix broken pages, slow load times, and mobile rendering issues. Check robots.txt for AI crawler blocks. If your site isn't technically sound, nothing else matters.
Rewrite your top 10 pages for direct answers. Take your highest-traffic pages and restructure them. Move the answer to the first two sentences under each heading. Add an FAQ section. Make sure every H2 promises a specific answer and delivers it immediately.
Build your entity authority. Get listed on industry review sites. Earn mentions in reputable publications. Participate authentically in Reddit and Quora threads where your expertise is relevant. AI tools notice when your brand appears consistently across trusted sources.
Track AI citations. Set up monitoring for your brand name and target keywords across ChatGPT, Gemini, and Perplexity. Lookelo's AI citation tracker shows you which queries trigger mentions and where you're missing opportunities.
Repeat for your competitors' keywords. Find the queries where competitors are getting cited and you're not. Analyze the pages they're citing. Structure your content to answer those same questions better — more directly, more authoritatively, with more specific data.
FAQ
Is SEO still relevant in 2026 with AI search tools?
Yes. AI search tools like ChatGPT, Gemini, and Perplexity rely on traditional SEO signals — crawlability, content quality, and authority — to find and cite web pages. SEO is the foundation AI search is built on.
What is the difference between SEO and GEO?
SEO focuses on ranking in traditional search results. GEO (Generative Engine Optimization) focuses on being cited in AI-generated answers. Google says GEO is still SEO, just applied to a new surface. The underlying practices are the same: clear structure, direct answers, and strong authority.
How does ChatGPT decide which websites to cite?
ChatGPT uses Bing's search index and evaluates content for relevance, clarity, and authority. It favors pages that answer questions directly, use clear heading structures, and come from sources with strong third-party credibility signals.
Do I need special schema markup for AI search?
No. Google's official documentation says special schema is not required for AI features. Standard schema types — FAQ, Article, Product, LocalBusiness — are still helpful for helping all crawlers understand your content.
How can I check if my website is being cited by AI tools?
Use an AI citation monitoring tool like Lookelo to track brand mentions across ChatGPT, Gemini, Perplexity, and Google AI Overviews. You can also check Google Search Console for AI-referred traffic patterns.
Will AI search reduce my website traffic?
It can. AI Overviews and zero-click answers reduce organic CTR — by as much as 65% for queries where AI Overviews appear and your site isn't cited. But sites that are cited in AI answers see 35% more organic clicks than uncited competitors at the same ranking position. The goal is to be the source, not just the link.
What is Answer Engine Optimization?
Answer Engine Optimization (AEO) is the practice of structuring content so AI-powered answer engines — like ChatGPT, Google AI Overviews, and Perplexity — can extract and cite your information directly. It emphasizes direct answers, FAQ structures, and entity authority.
Do I need to create llms.txt files for AI search?
No. Google's AI optimization guide explicitly says llms.txt files are not required for Google Search. Some AI platforms may use them, but they are not a ranking or citation factor for Google's AI features.
Your SEO Didn't Break. It Just Needs to Show Up in More Places.
AI search didn't kill SEO. It made SEO more important — and more interesting. The boring stuff still matters: fast pages, clean code, good content, real authority. But now there's a second layer: structuring what you write so AI tools can extract it, cite it, and send curious readers your way.
The brands winning in 2026 aren't the ones with the biggest SEO budgets. They're the ones who adapted fastest — who realized that the same content that ranks on Google can also get cited by ChatGPT, referenced by Perplexity, and surfaced in Gemini, if it's built right.
You already have the foundation. The question is whether you're optimizing for the search that actually happens now.