Your buyers stopped scrolling.
They ask, and get one answer.

It is the work of getting your company accurately described on the sources AI assistants read, so that when somebody asks about your category the answer has a reason to mention you. It is sometimes called generative engine optimization, GEO, or answer engine optimization. The names differ, the job does not.

Classic SEO competes for a position on a page of links, where being tenth still gets a click. An assistant returns one answer naming a handful of companies, so there is no tenth place. The other difference is where the signal lives: a search engine reads your site to rank your site, while an assistant reads everyone else's pages to decide who to mention.

No, and you should not buy from anyone who says otherwise. We do not guarantee a position in ChatGPT or any other model. Nobody can. We measure where you stand, and we fix the layer you control. What we guarantee is the work itself: Your story publishes on all 7 outlets, or you get your money back.

Run the free check. It takes about a minute, asks real questions from your category to live models, and shows you the unedited answer with your name highlighted if it is there and the competitors named instead if it is not. No account, no card.

Publication happens in days. The models picking it up is slower and depends on how often they recrawl your category, which is why we re-run the same questions rather than asking you to take our word for it. Anyone quoting you a precise date for a model update is guessing.

Not remotely, and this is the single most useful thing in our data. Across the answers we have collected, the average number of sources a model reads before answering ranges from 1.5 to 52.1 depending on which one you ask, and the rate at which a model names the company being checked varies by roughly a factor of two between the most and least generous. Optimising for one model and assuming the rest follow is a mistake.

No, and the data is blunt about it. Across every answer we have collected, the 20 most-cited domains account for 6% of all citations, spread across 17,713 distinct sites. There is no short list to buy your way onto. Breadth of accurate description beats a handful of placements.

The check is free. A full measurement across 50 buyer questions and every model is $97. Getting described on high-authority outlets starts at $999 for a bundle, or per outlet from the catalogue if you would rather start with one.

It finds the questions your buyers ask AI assistants, records what each assistant answers and which pages it reads, then works to get your company described accurately on those pages and on your own site, and asks the same questions again. The labels vary (AI SEO, GEO, AEO, LLM SEO); the work is the same.

Tool subscriptions we read on their own pages on 3 Nov 2026 start at $29 a month (Otterly.AI Lite) and $95 (Peec AI Starter); the agency pages we read published no package price. At Essentras the check is free, the AI Visibility Baseline is $97 once, the Tracker is $49 a month, and the work itself is quoted after the free check.

In-house works when someone can own 3 jobs: tracking the same questions every week, fixing your own pages, and earning or buying accurate descriptions on third-party sites. Hire when you lack the time or the outlet relationships. Either way, start with a baseline, so you can tell whether anything moved.

Measure them apart. Google: clicks, impressions and average position in Search Console, where AI Overviews and AI Mode count inside the Web search type. Assistants: the same questions asked again on a schedule, counting answers that name you (mentions) and answers that cite your site (citations), assistant by assistant.

Google says there are no additional requirements to appear in AI Overviews or AI Mode, and no special optimizations are necessary. What changes is the measurement: Search Console reports those appearances inside the Web search type, not in a separate AI report.

Nobody can promise that. A placement puts an accurate, dated description of your company on an outlet the assistants may read; whether a model uses it is the model’s decision. A placement is paid and should be labeled as paid. Measure before and after with the same questions.