Your customers are asking ChatGPT who to hire. Find out whether the answer includes you. We send five real buying questions to ChatGPT, Perplexity, Google AI Overviews and Gemini, then show you every answer, whether your brand was named, whether your site was cited, and exactly which competitors were handed the recommendation instead.
Use the words a customer would use, not your internal service name.
Leave it blank if you sell nationally.
Your business name, your website, and what you sell in the words a customer would use.
Five real questions a customer would type, sent live to every answer engine we can reach.
Every answer, scored for whether you were named, whether you were cited, and who beat you.
Every result is a live call to that engine. AI answers are probabilistic, so wording shifts between runs. Read the pattern across the five questions rather than any single line.
For twenty years, search was a list. You optimized for a position on that list, and the click was yours to win. An AI assistant does not hand back a list. It hands back a decision: two or three names, a sentence of reasoning, and sometimes a link. Most people act on that shortlist without scrolling anywhere. If your business is not one of the names, the ranking you worked years for never enters the conversation.
This is the gap the scan measures. It is entirely normal to rank on page one for your main keyword and still be invisible inside the answer, because the two systems select differently. A search engine ranks documents. A language model assembles an answer from whatever it can retrieve, parse and attribute with confidence, which in practice means roundup articles, review platforms, forum threads, structured profile data and source pages written to be quoted.
Each row is one engine answering one question. Named and linked is the outcome you want: the assistant used your brand and attributed your domain as a source. Named, not linked means the model knows you exist but pulled the fact from somewhere other than your site, usually a directory or a roundup, which means someone else owns the citation that should be yours. Not named means that for a buyer asking that question, you do not exist.
The pattern across engines matters more than any single row. Showing up in Google AI Overviews but nowhere else usually means your classic SEO is carrying you and your entity signals are thin. Showing up in Perplexity but not ChatGPT often means you have recent third-party coverage but little that the larger models have absorbed. Showing up nowhere, including on the final question that names you directly, means the models have no confident record of your business at all, which is the most fixable version of this problem.
Three levers, in order of impact. First, entity clarity: the model has to be certain who you are, what you sell and where you operate. That comes from consistent naming, Organization and Service schema, a sameAs graph tying your profiles together, and a site that states the basics in plain text rather than implying them through design. Second, citable source pages: content built so a specific passage answers a specific question cleanly enough to be lifted and attributed. Third, third-party presence, which is the one most brands underinvest in, because assistants lean heavily on the roundups, directories and community threads that already rank for the category question.
None of that replaces conventional SEO, it compounds on top of it. The sites getting cited are almost always the ones that were already technically sound, then made themselves legible to a model. See how our AI SEO service builds all three.
Spend the first scan on your highest-margin service, described the way a customer would describe it rather than the way your website names it. Spend the second on a secondary service or a second market, since visibility rarely transfers between them. Spend the third on your strongest competitor: seeing which of their pages the models cite tells you precisely what a citable source page looks like in your category.
If you want to go further first, our free map rank checker covers the same question for Google Maps, and the free AI SEO training walks through structuring content that AI systems can actually quote.
Check your own data
Google used to fold AI traffic into the standard Search report with no way to separate it. That changed. Search Console now has a dedicated Generative AI performance report, announced on 3 June 2026 and rolled out to every site worldwide by 31 August 2026. If you have not looked at it yet, it is the first place to go after running the scan above.
Open Search Console for the property you want
You need verified ownership of the domain. If you have never verified it, that is step zero and it takes about five minutes with a DNS record.
Go to All reports and tools, then Performance reports
The generative AI reports sit alongside the familiar Search results and Discover reports rather than inside them.
Open Generative AI performance report (Search)
This covers AI Overviews and AI Mode together. There is a separate Discover version for generative features inside Google Discover.
Filter by page, country, device and date
Date granularity runs from daily to monthly, in Pacific Time. Dotted lines mean the data is still preliminary and will move.
Impressions only
No clicks, no click-through rate, no average position. You can see that you appeared, not what it was worth.
No queries or prompts
There is no query dimension at all, so you cannot see what the user actually asked to make you appear.
Google only
AI Overviews and AI Mode. ChatGPT, Perplexity and Claude are not in Search Console and never will be.
Your site only
You cannot see a competitor being cited where you are not, which is the comparison that actually matters.
No API
The generative AI data is not in the Search Analytics API or the BigQuery export, so it cannot be pulled into a dashboard yet.
Search Console tells you that you appeared. The scan above tells you what was asked, what the answer said, whether it named you, and which competitors it named instead. Run both.
Open Search Console, go to All reports and tools, then Performance reports, then Generative AI performance report (Search). Google launched it on 3 June 2026 and finished rolling it out to all sites on 31 August 2026. It covers AI Overviews and AI Mode together and can be filtered by page, country, device and date. Be aware of what it leaves out: it reports impressions only, with no clicks, no click-through rate, no average position and no query data, and it is not available through the Search Analytics API or the BigQuery export yet.
No. Search Console only reports on Google surfaces, so its generative AI report covers AI Overviews and AI Mode and nothing else. ChatGPT, Perplexity and Claude publish no equivalent reporting at all. The only way to know whether those assistants name your brand is to ask them the questions your buyers ask and read the answers, which is what the scan on this page does.
It is a check of whether AI assistants name your business when someone asks them a buying question. Instead of looking at where you rank on a page of blue links, the scan asks ChatGPT, Perplexity, Google AI Overviews and Gemini the questions your customers actually ask, then reports whether your brand appeared in the answer and whether your domain was cited as a source.
Because an AI answer replaces the click. When a buyer asks an assistant who the best provider is, they get a short list of two or three names and usually act on it. Traditional rankings feed that list but do not guarantee a place on it: assistants pull heavily from roundups, review sites, forums and structured source pages. Plenty of businesses that rank on page one of Google are never named in the answer at all.
They largely describe the same work under different names. GEO stands for generative engine optimization, AEO for answer engine optimization, and AI SEO is the plain-language umbrella term. All three mean making your business the source an AI model reaches for: clear entity signals, pages written to be quoted, structured data the model can parse, and third-party mentions the model already trusts.
Every result is a live call to that engine, not a cached estimate, so what you see is what the engine returned at that moment. AI answers are probabilistic, so the exact wording changes between runs. The pattern is what matters: if four of five questions come back without your name, that is a real gap, not noise. Run it again in a week and compare.
Three things, in rough order of impact. Entity clarity, so the model knows who you are, what you do and where, through consistent naming, schema markup and matching profiles across the web. Citable source pages, meaning content that answers a specific question in a way that can be lifted and attributed. And third-party presence, because assistants lean on roundups, directories, review platforms and forum threads far more than most brands expect.
Yes. You get three free scans, which is enough to test your main service, a second service line and a competitor. There is no card and no trial. We built it because we run this analysis for our own clients, and it makes the conversation about what is broken much shorter.
A scan full of red is a diagnosis, not a fix. This is the work that changes it.
Entity markup, citable source pages, and the third-party presence AI models already trust. The work that gets you named.
Explore AI SEOA 25-minute walkthrough of structuring content so Google, Bing and AI assistants can quote it. Template included.
Watch the trainingThe same question for Google Maps: a 49-point heatmap of where you actually rank across your service area.
Run a map scanBook a free audit call. We will go through your scan line by line, show you which sources the models are pulling from instead of yours, and lay out what it takes to replace them.