Learn GEO, then go get cited.
Guides, definitions and original data on how AI search works and how to get your brand named in the answers buyers now trust.
Reading up because AI ignores your brand? Run the free visibility check →
What gets a brand cited by AI: what we have measured, and what we have not.
Our studies, each labelled by engine, collection route and date. Nothing gated, no unmeasured numbers.
Read the report →Latest guides and news
First on Google does not guarantee a mention in ChatGPT
Discoverable compared 1,000 buying questions across six consumer AI services, collecting 17,537 completed AI answers and 3,000 matching Google results captures. ChatGPT did not name the business at Google's first organic position in 72.5% of eligible answers.
Read →Why has ChatGPT stopped recommending my brand?
A diagnostic ladder for a brand that used to be named in ChatGPT answers and now is not: how to tell a real drop from sampling noise, then the causes to check in order, from your own account settings to a retrieval change at OpenAI.
Read →Why does AI recommend my competitors and not me?
AI names your competitors because the pages it cites name them. A per-rival diagnosis you can run prompt by prompt: read the cited sources, compare depth, wording, review recency and the sub-questions the rival owns, then score the gap so you know what to fix first.
Read →How to find the sources AI uses to recommend your competitors
The procedure for building the list of pages that currently decide who AI recommends in your category: where citations show on each engine, why only cited sources count, a ten-prompt worksheet with the product's own access classes, and what to do with the list once you have it.
Read →How many prompts should you track for AI visibility?
Five to ten prompts by hand is a diagnosis; 25 fixed prompts checked weekly is the smallest set that gives a trend you can defend against sampling noise; 200 covers one product's buying journey; 1,000 per brand is for product lines and markets. One page that reconciles every prompt count on this site.
Read →ChatGPT has stopped citing Reddit
Reddit's share of ChatGPT Search citations fell by about 86 per cent in the first half of August 2026, from roughly 3.8 per cent of cited links to 0.5 per cent. The cause has not been confirmed, and the lesson is about concentration risk rather than Reddit.
Read →ChatGPT ads arrive in the UK
OpenAI's help centre lists ads in ChatGPT as live in the UK from 11 August 2026, alongside Brazil, Japan, Mexico and South Korea. The unit sits under the answer with a Sponsored label, and OpenAI says buying one does not change whether the answer above it names your brand.
Read →AI shopping agents just won the right to browse Amazon
On 4 August a US appeals court ruled Amazon cannot use the main federal anti-hacking law to block Perplexity's Comet shopping agent, finding it is the user, not the AI firm, who accesses Amazon. Agentic commerce just got a clearer legal runway, which raises the stakes on being the brand an agent picks.
Read →Google puts image generation inside AI Overviews
Google announced on 14 July that AI Overviews will generate images from a text prompt rather than only surfacing photographs from real websites. A drawn image has no page behind it and nothing to click, which removes another route by which brand imagery reached searchers.
Read →AI visibility for startups: how a new brand gets named by AI
AI has never heard of your startup, which means retrieval is the only way in. A step-by-step playbook: entity page, cited listicles, honest reviews, communities, and how to measure progress from a zero baseline.
Read →How to get your brand into Google AI Overviews
To get your brand into Google AI Overviews, rank for the full cluster of questions behind your target query, open each matching page section with a direct answer, keep your entity data consistent across the web, and earn a place on the third-party pages the Overview already cites.
Read →How to measure AI share of voice for your brand
A measurement guide marketers can hand to their exec: define AI share of voice, build a prompt set from real buyer questions, run it weekly, and report the three numbers that matter.
Read →Does your brand need a Wikipedia page to be visible to AI?
No. A Wikipedia page helps the answers assistants give from memory, but it is not required for visibility, and most SMBs will not meet notability anyway. Here is how assistants use Wikipedia and Wikidata, and what to build instead.
Read →Schema markup for AI search: what actually matters
Schema markup helps AI search mainly through entity disambiguation: Organization and Product markup feed the knowledge graphs that ground AI answers, while rendered text does the retrieval work. Covers a priority order for schema.org types, the ones to skip, and where llms.txt stands as of mid 2026.
Read →How to Fix Wrong Information AI Gives About Your Brand
A practical guide to correcting the wrong claims AI assistants make about your brand, from stale pricing to discontinued products, with each fix ranked by leverage and given a realistic timescale.
Read →Do AI assistants read your reviews? G2, Trustpilot and Reddit in AI answers
AI assistants fetch and quote review pages when they answer buying questions. This guide covers which review surfaces actually get cited and how to earn the kind of written, specific reviews that end up quoted in AI answers.
Read →The 30-minute AI visibility audit (checklist)
A six-step do-it-yourself audit of how AI assistants see your brand, with exact instructions for each check and a scoring method you can repeat every month.
Read →Why doesn't ChatGPT mention my brand?
The five most common reasons ChatGPT leaves your brand out of its answers, in order of likelihood, with a self-check and a fix for each.
Read →Does Reddit improve AI visibility? What actually works
Reddit heavily influences what AI assistants say about brands. This guide explains which threads engines cite and why genuine participation beats manufactured mentions, which can poison your record for years.
Read →llms.txt: what it is and whether you need one
A guide to the llms.txt proposal: the honest state of AI engine adoption, and how to write a file that doubles as a useful content audit.
Read →Google's AI Mode Now Fills Shopping Carts: What It Means for Brand Visibility
Google's Connected Apps launch on 16 July lets AI Mode add groceries to an Instacart cart and act inside Canva and YouTube Music without leaving search. Brands now compete to be named in answers that act, not just inform.
Read →How AI assistants choose which brands to recommend
AI assistants recommend brands by finding consensus across independent sources, whether they answer from training memory or live web retrieval. This guide explains both modes, why third-party pages outweigh your own site, how the major engines differ, and what to do about it in priority order.
Read →Is my brand visible in AI search? How to find out in an hour
A hands-on, one-hour audit for checking whether AI assistants name your brand: ten buyer queries, three assistants, five things to record per answer, and how to read the three patterns the results almost always fall into.
Read →The GEO content checklist: pages AI actually cites
Concrete on-page guidance for earning AI citations: answer-first openings, one claim per sentence, question-shaped headings, visible dates, extractable structure and entity clarity, finishing with a printable 10-point checklist.
Read →We asked AI about 30 brands. 27 were described wrongly
Our July 2026 test asked an AI assistant about 30 real brands, first from memory and then with live web search. From memory, 27 of 30 were described with at least one materially false claim. Live search cut that to 13. Here is the method, the results and what brands can do.
Read →How does Perplexity choose which sources to cite?
Perplexity runs a live web search for every answer and cites the pages its model actually uses. This guide covers the observable selection patterns (answer-shaped pages, freshness, domain trust, clean extractable statements) and the concrete steps to earn citations this week.
Read →How to check if your brand appears in Google AI Overviews
Google AI Overviews are generated per query, per user and per location, so a single search tells you very little. This guide walks through checking your brand manually this week: which queries to run, how to spot citations, what to do if you're absent, and where tooling takes over.
Read →Start here
The best GEO tools in 2026
Honest comparison with verified pricing, rivals included and credited where they win. Updated 20 July 2026.
DataThe AI Visibility Index: 40 brands measured
Original data: how visible 40 household brands are in AI answers. Fame doesn't guarantee visibility.
DataGEO statistics 2026
The measured numbers behind AI search visibility, first-party and free to cite. Fame does not equal visibility.
Pillar guideThe GEO playbook: how to get cited by AI in 2026
The full pillar guide to measuring and improving your presence in AI answers.
GuideAnswer-first writing for AI citations
How to structure a page so an AI engine can lift and cite one clean sentence.
GuideFind the sources AI cites for your competitors
Where AI pulls category answers from, how to list those pages for your category, and how to get included on them.
Glossary
See where you stand in AI search, and fix it.
Run a free check in under a minute. No card, no login to see your score.