Visibility
Through the same groundwork: content that answers real questions, technology that makes it readable for machines, and a brand that gets mentioned beyond your own site. Here is what matters in 2026, plus ten concrete tips you can act on today.
Last updated: 8 July 2026
The terms
Three acronyms, one goal: being the answer when someone asks.
Showing up in the classic search results on Google and Bing. The technology, content and authority that make your pages rank when customers search.
Showing up when the search engine answers the question directly, for example in Google AI Overviews and AI Mode, where a growing share of searches end without a single click.
Getting mentioned and cited when AI assistants like ChatGPT, Perplexity, Gemini and Claude compose their answers. This is what we call AI search optimization.
Why now?
The question is whose name they get as the answer.
Note: these figures come from third-party data and Google's own statements. AI search is evolving quickly, so treat them as direction rather than exact values. This page reflects the state of knowledge in June 2026.
Free audit
Our AI Visibility Audit measures how visible your brand is in Google's AI answers, how AI describes you and whether your domain is cited as a source, compared with your competitors. You get a score, a competitive picture and a prioritized action list within one working day. Free and with no obligations.
Request a free audit →The foundation
Google states explicitly that no special AI optimizations are required: AI Overviews and AI Mode are built on the same ranking and quality systems as regular search (Google Search Central, May 2025). Classic SEO is the foundation for AI visibility too.
In practice that means three things: technology that makes your site easy to crawl and index, content that matches what your customers actually search for, and authority in the form of links and mentions from other sites.
What has changed is how the answers are presented, and what it takes to get cited in them. That is where the tips below come in.
Playbook
Would you rather have the tips packaged as a browsable presentation, with all the statistics and sources collected in one place and ready to share internally? Enter your email and company for instant access to the playbook.
AI assistants can only cite what they can read. Many sites block AI crawlers in robots.txt without knowing it, often through old rules that block all bots.
Do this: open your robots.txt and make sure the settings are right: which bots are allowed (and not allowed) to visit which pages.
What about llms.txt? It does no harm, but Google's official guidance (May 2026) says it is not used for Search, AI Overviews or AI Mode. Focus on robots.txt instead.
LLMs favor content that gives a clear, complete answer early on the page. Long introductions make content harder to cite, and industry data suggests AI citations are drawn disproportionately from the first part of a page.
Do this: open every important page with two or three sentences that answer its main question directly, then go deeper.
The first major academic study in the field tested nine tactics across 10,000 queries. Three stood out: adding statistics, expert quotes and source citations increased visibility in AI answers by 30 to 40 percent.
Do this: review your most important pages and add concrete numbers, quotes from named experts and linked sources where it makes sense.
Source: Aggarwal et al., "GEO: Generative Engine Optimization", KDD 2024 (Princeton et al.)
A clear heading hierarchy, bullet lists, tables and FAQ sections make content easy to extract. Important: several AI crawlers do not render JavaScript, so content that only appears after JS rendering may never be read. Microsoft's official guidance points the same way: do not hide key content in tabs, PDFs or images only.
Do this: make sure your key content is in the HTML at page load. Test it: view the page source and search for your key text.
Source: Microsoft/Bing, "Optimizing your content for inclusion in AI search answers" (Oct 2025)
When someone asks a question, AI assistants perform a so-called fan-out: the question is split into several narrower sub-queries that are answered separately. A single search can trigger anywhere from a handful to twenty or so sub-queries. A cluster of focused pages covering the topic broadly gives you more chances to catch one of them.
Do this: map the questions your customers ask within your topic and cover them with a cluster of focused pages. One question per page usually beats one giant page about everything.
Fan-out is Google's own term for how AI Mode fans a question out into several sub-queries.
AI models learn about brands from the whole web: industry media, podcasts, forums, YouTube, LinkedIn and review sites. In Ahrefs' study of 75,000 brands, brand mentions were the factor that correlated most strongly with AI visibility, more than three times stronger than backlinks. Correlation is not causation, but the signal is clear.
Do this: ask ChatGPT and Perplexity which sources they cite in your industry. Then make sure you appear in the ones that matter most.
Source: Ahrefs, study of 75,000 brands (2025)
Credibility carries weight in both worlds. Named authors with a role, credentials and a real author page give LLMs a reason to trust your content, and to cite it.
Do this: where possible, give every important article a byline with name and title, linked to an author page with background and expertise. Show publication and update dates.
LLMs prioritize recently updated content. In Ahrefs' analysis of 17 million AI citations, the content AI cites was on average about 26 percent fresher than regular search results.
Do this: set up a quarterly routine to update your content (text, figures, years, examples and so on) on your most important pages. Ideally, display the update date clearly.
Source: Ahrefs, 17 million AI citations (July 2025)
AI avoids citing generic content that exists on a hundred other sites. Your own data, such as a survey, a benchmark or statistics from your business, is the most citable thing you can publish. The industry calls this information gain. Whoever owns the number owns the citation.
Do this: publish an annual survey or an industry benchmark built on your own data. Give it a name people can refer to.
AI traffic is still small in volume, but several independent measurements suggest it often converts better than average organic traffic. What you do not measure, you cannot improve.
Do this: ask ChatGPT, Perplexity and Gemini your most important customer questions once a month and note whether and how you are mentioned. Segment AI referral traffic in GA4.
Note: AI answers vary between people, sessions and phrasings. Treat your own tests as an indication, not an exact measurement.
Content
Without content there is nothing for Google to rank or for an AI to cite. Visibility work and content work belong together: search data and AI questions reveal what your customers are asking, content answers it, and structure makes those answers citable.
We help you prioritize which content makes the biggest difference for your visibility, and how to shape it so it can be read, ranked and cited.
About SEO, AEO and GEO.
GEO (generative engine optimization) is about getting mentioned and cited when AI assistants like ChatGPT, Perplexity and Gemini compose their answers. AEO (answer engine optimization) is about showing up when the search engine answers the question directly, for example in Google AI Overviews. The terms overlap, and both build on the same foundation as classic SEO.
No. Google states explicitly that AI Overviews and AI Mode run on the same ranking systems as regular search. Good SEO is the foundation, and GEO adds what increases your chances of being cited: clear answers, sources, structure and a brand that gets mentioned beyond your own site.
No, and be skeptical of anyone who promises that. AI answers vary between people, sessions and phrasings. What can be done is to improve your odds and measure the development month by month.
It depends on your starting point. Technical blockers, like AI crawlers being shut out, can be fixed right away, while content and brand work usually needs a few months before it shows in the measurements. That is why we measure from day one.
Send a message through the contact form at the bottom of this page. We look at how you show up today and suggest a sensible first step.
Explore
It all starts with a hello. Tell us a bit about your project or challenge, and we'll get back to you.
