“what is the best pizza place in Austin?”
Someone asks an AI that today. We measure where your business shows up, in AI answers built from web search and on the Google map those AIs read, and what changes when it does not.
It runs in front of you, question by question. No card.
Asking one AI only answers for one AI.
Across 100,000 prompts, only 11 in every 100 cited domains showed up in ChatGPT and Perplexity at the same time. Ranking in one engine says nothing about the other.
The subscription measures three engines. The free audit measures one, and says on screen which ones it left out.
11 in every 100
cited domains showed up in both engines. The other 89 showed up in just one.
Profound, 100,000 prompts.
Does your brand show up when the AI answers?
A free, live audit: we show where you stand on the Google map and, question by question, whether your brand shows up in ChatGPT answers via API with web search. Each person’s app also uses the map, their memory and their location.
The full audit, with the other engines and every question in your category, comes later and with us: it is not what this button runs.
The account you check from is the one that errs in your favor.
Almost every owner has typed their own name into an AI to see what comes out. The trouble is that the answer they see is not the answer a stranger sees: their account carries history, and history pushes the answer toward what pleases them.
Measuring takes a clean account, the same question text repeated identically across clients, and a record of which version of which engine answered. None of that fits in a tab opened in a hurry.
- Memory
- What you asked before keeps shaping what the AI answers later.
- Personalization
- The answer adapts to whoever is asking, and you are the person with the most signals about your own business.
- A/B testing
- Providers serve different versions to different audiences. Two people asking the same thing get different answers, with no notice.
Measuring is a commodity. What you buy is the fix.
Dozens of tools will tell you whether your brand showed up. What is missing is what to do when it does not, ready to paste and ready to act on.
1
The structured data block, ready to paste
We generate JSON-LD typed for your category, filled in with what your site already publishes.
Anything we do not know ships with a visible marker, never a plausible value: a made up address passes validation and misinforms, which is worse than an empty field.
2
A brief for every question you lost
For each question where your brand did not appear, we list who did, where the AI took the answer from, and what your content needs to carry.
The requirements come from levers a study actually measured. Keyword density is left out on purpose: it is the one strategy the study measured as null.
3
Where you need to be, with the sign-up link
We check page by page the sources the AI cited and tell you which ones your business is missing from, with the exact address to fix it.
In local services, most of what an AI cites is not the brand's own site. Fixing only your page solves the smaller term.
The most cited source in your sector is the one you control.
In the largest published local study, the Google profile accounts for 28.63% of all local citations. Yelp sits at 9.53% and Facebook at 2.23%.
Claiming and completing that profile costs nothing, and it carries the highest effect per hour of work in the whole registry. It ships in the report with the study reference beside it, because impact without provenance is indistinguishable from a number someone made up.
- Google profile
- 28.63%
- Yelp
- 9.53%
- 2.23%
- All other sources
- 59.61%
The price is here, not behind a form.
One price, in two moments: setup builds what is missing, the monthly fee keeps it and measures. The audit that opens the conversation is free.
See what is included- R$ 1,490once, at the start
- We build what is missing: Google Business Profile, profiles on the directories the AIs read, JSON-LD on the site, service pages, consistent name, address and phone. That day's audit is the baseline.
- R$ 990/monthper month, 6 months minimum
- We keep what was built, run the review routine with you and measure again every week. Every month, a report of what we did and what changed.
What this product does not know.
Every tool in this category has the same instrument limits, and almost none of them publish theirs. These are ours, worded the way they appear in your report.
- The API is not the app
- We measure through the engines' APIs. The answer your customer sees in the app goes through memory and personalization, and can differ from ours.
- One measurement per question
- Each question runs once per engine. A St. Gallen study measured a standard error of 0.370 for an isolated run, which is close to uninformative on its own. Every competitor runs it the same way, and none publish it.
- The city goes inside the question
- We put the city name inside the question text. We measure what the AI answers about your city, not what it would answer from inside it.
When a round measures less than it should, the report changes voice, says it is not conclusive, and the only recommendation it makes is to measure again. Prescribing a fix for an absence nobody measured sends you spending in the wrong place.