Adamiro AI Visibility · Answer engine tracking
Your buyers asked an AI about your category this morning. You have no idea what it said.
AI Visibility runs the questions your buyers actually ask against ChatGPT, Perplexity and Google AI Mode on a schedule, and reports what came back: whether you were named, whether the answer chose you, who was named instead, and which pages the answer was built from.
What changed
Search used to send people to you. Now it answers for you.
Someone evaluating your category types a question into ChatGPT. They get a paragraph with three or four companies named in it, and a handful of links underneath. If your name is in that paragraph, you are on the shortlist. If it isn’t, you were never in the running, and nothing in your analytics will ever tell you it happened.
This is not a ranking you can look up. The answer is generated fresh each time, it differs by how the question is phrased, and it draws on pages across the web rather than only your own. The only way to know where you stand is to ask the same questions repeatedly and record what comes back.
You cannot fix a shortlist you can’t see.
TL;DR
How it works
Ask the buyer’s questions. Record every answer.
1 · The questions
Adamiro writes the questions a buyer would ask about your category, from what it already knows about your business. They range from someone who has never heard of you to someone comparing two shortlisted vendors. On Business and above you can write your own.
2 · The engines
Every question is put to ChatGPT, Perplexity and Google AI Mode as a real buyer would put it, on the schedule you choose. Each answer is stored in full, so a claim about your visibility can always be traced back to the words an engine actually produced.
3 · The reading
Each answer is checked for whether you were named, how early, in what tone, who else appeared, and whether the answer put one company forward as its pick. That becomes a presence grade weighted toward the questions asked closest to a buying decision, because being named in a comparison matters more than being named in a definition.
4 · The work
The questions where you are missing become a list of things to write, each one sent to Voice as a draft with the gap it is meant to close. Publish through Reach and the next report tells you whether the answer changed.
Sources
See which pages the answer was built from.
Perplexity and Google AI Mode publish the pages behind their answers. AI Visibility collects them and sorts them by what kind of page each one is: your own pages, a competitor’s, a discussion thread, a social channel, a shop, or everywhere else. The kind tells you what getting cited would take. A thread wants a reply, a shop wants a listing, a trade title wants a quote.
A domain that keeps getting cited in answers where you are absent is something specific you can go and do: a page you should be on, a comparison you should be listed in, a thread worth a considered reply. It beats guessing at keywords.
It also checks whether the assistants can find you as a company at all: a Wikidata item that points at your site, a Google entity, Organisation schema on your home page, named authors on the pages that get cited, a G2 or Capterra listing, a roundup page that names you. Each missing one comes with instructions, and none of it spends a task.
ChatGPT does not publish its sources. The dashboard says so on the page rather than quietly reporting a smaller number.
In the dashboard
What a report tells you.
AI Presence Grade
One letter, A to F, weighted toward decision-stage questions. It moves when something real changes, not when an engine rephrases itself.
Answer position
Where in the answer your name appears. Being the first company named reads entirely differently from being the fourth, and buyers stop reading.
Share of voice
How often you’re named against the competitors you nominate. We only count questions that don’t already say your name, so asking about you by name can’t inflate the score. Name your competitors yourself, so you’re measured against who you actually lose deals to.
Visibility by buying stage
Strong on “what is” questions and absent on “best for” questions is a common and expensive pattern. Split out so you can see which end you are losing.
Cited sources
Every page an engine drew on, grouped by domain and sorted by what kind of page it is, so you can see what getting onto it would take. Available from Growth up.
What it is not
Two things worth being straight about.
This is a measurement, not a lever. Nobody can promise an AI engine will name your company, and anyone who does is selling something they cannot deliver. What AI Visibility gives you is a repeatable reading and a list of gaps to close, so the work you do afterwards is aimed at something real.
And answers vary. The same question asked twice can come back differently, which is exactly why one spot-check tells you nothing and a trend over weeks tells you a great deal. Reports run on a schedule for that reason.
AI Visibility by plan
| Feature | What it delivers | Solo | Team |
|---|---|---|---|
| Enabled | ✓ | ✓ | |
| Prompts | Unlimited | Unlimited | |
| Engines | Openai, Perplexity, Gemini | Openai, Perplexity, Gemini | |
| Cadence | Weekly | Weekly | |
| Custom Prompts | ✓ | ✓ | |
| Citations | ✓ | ✓ | |
| Audience Breakdown | ✓ | ✓ | |
| History Months | Unlimited | Unlimited |
What AI engines tell your buyers about you
AI Visibility starts from the questions a buyer would type, not from keywords.
Adamiro writes them from what it already knows about your business, tagged by buying stage. They range from someone who has never heard of you to someone comparing two shortlisted vendors. You can add your own; the first ten a week are in the seat, then one task per ten.
Read the full demo, step by step
- The questions your buyers actually ask — AI Visibility starts from the questions a buyer would type, not from keywords. Adamiro writes them from what it already knows about your business, tagged by buying stage. They range from someone who has never heard of you to someone comparing two shortlisted vendors. You can add your own; the first ten a week are in the seat, then one task per ten. (Behind the scenes — reads Profile: The question set is generated from your company profile, so it describes your category rather than a generic one.)
- Asked on a schedule, recorded in full — Every question is put to ChatGPT, Perplexity and Google AI Mode, and every answer is kept. One reading tells you nothing, because the same question can come back differently. A trend over weeks tells you a great deal. Each report scores whether you were named, how early, in what tone, and who appeared instead, and splits the result by buying stage.
- Twelve readings, not one — Every report is kept in full: the grade it earned, what each engine said, and the recommendations that came out of it. A single reading is a mood. What you want is the shape of the line: whether the questions closest to a sale are starting to name you, and which week that changed. Any report can be reopened exactly as it was written.
- One grade, weighted where it counts — The AI Presence Grade leans on decision-stage questions, because being named in a comparison matters more than being named in a definition. Answer position sits beside it: appearing first in an answer reads very differently from appearing fourth, and buyers stop reading. Visibility is split by buying stage so you can see which end you are losing.
- See which pages the answer was built from — Perplexity and Google AI Mode publish their sources. AI Visibility collects them and sorts them into your pages, your competitors' pages, and everyone else's. That last group is usually the interesting one: the review sites, forum threads and directory listings shaping your category. A domain cited again and again in answers where you are absent is something specific you can go and do. ChatGPT does not publish its sources, and the tab says so rather than quietly reporting a smaller number.
- Measured against the competitors you name — You nominate the companies you actually lose deals to, and share of voice is measured against them, counted only on the questions that don't already say your name. A question that names you is testing how an engine describes you, not whether it can find you at all, so those are left out of this number. Add a domain alongside the name and Citations can also tell you when an answer was built from that competitor's own site.
- Gaps become things to write — The questions you are missing from arrive as recommendations and content ideas, each one ready to hand to Voice. This is where a report turns into work. Publish through Reach, and the next report tells you whether the answer changed. (Behind the scenes — writes Voice ideas: Send to Voice carries the gap with it, so the draft knows which question it is meant to answer.)
Questions about AI Visibility
What is AI Visibility?
AI Visibility tracks what AI answer engines say about your company. It runs a set of buyer-realistic questions against ChatGPT, Perplexity and Google AI Mode on a schedule, then reports whether you were named, how early in the answer, in what tone, whether the answer chose one company and whether that was you, which competitors came up instead, and which web pages the answer was built from.
If you need a simpler explanation you can find one here: AI Visibility Explained
Which AI engines does it check?
ChatGPT, Perplexity and Google AI Mode. Every seat runs a weekly read across all three engines, because a score built from one engine is not comparable with a score built from three. 10 prompts are included; extra prompts cost 1 task per 10, per weekly read.
Can you guarantee my company will be named?
No, and neither can anyone else. AI answers are generated fresh each time and no vendor controls what an engine says. What AI Visibility gives you is a repeatable measurement and a specific list of gaps to close, so the work you do afterwards is aimed at something real rather than guessed at.
Is this the same as SEO?
No. SEO is about ranking in a list of links. AI Visibility is about being named inside a generated answer, which draws on pages across the whole web rather than only your own site and never produces a ranking you can look up. The two overlap, since strong pages help both, but you can rank well and still be absent from the answer.
What are citations and why do they matter?
When an engine answers a question, it builds that answer from pages on the web. Perplexity and Google AI Mode publish which ones. AI Visibility collects them and sorts them by what kind of page each one is: yours, a competitor’s, a discussion thread, a social channel, a shop, or everywhere else. A site cited again and again in answers that never name you is the specific reason you are invisible for those questions. ChatGPT does not publish its sources, and the dashboard says so rather than quietly reporting a smaller number.
How often does it run, and what does it cost?
Every seat gets a weekly read, 10 prompts included. A complete report is one full run across every active question and every engine within that budget, and extra prompts beyond the 10 cost 1 task per extra 10. Regenerating your questions, managing competitors, editing your domains and reading any tab are all free.
What happens to the questions where I am missing?
They become work. Each report ends with recommendations and content ideas, and one click sends an idea to Voice as a draft brief carrying the question it targets. Publish through Reach and the next report reads the same question again, so you can see what moved.
Is being named by an AI the same as being chosen?
No. An answer can list five companies and recommend one, or recommend none. In a study of how coding agents pick developer tools, reported by The New Stack in September 2026, Armature counted PayPal named 139 times and chosen zero, and LangChain named 194 times and chosen four. Developer tools are a different market from yours. The pattern is the same one, and it is why your report counts the answers that chose you separately from the answers that only named you.
Is this AEO or GEO?
AEO and GEO are two names for one job. Answer engine optimization is earning a place in the answer an assistant gives. Generative engine optimization is changing what the model behind that assistant already believes about you, before anyone asks it anything. AI Visibility measures the first every week, and the recommendations it writes go after the second. We call the product AI Visibility because that is the phrase buyers use when they ask us for it.
Find out what they’re being told.
AI Visibility is included on every seat: a weekly read across all three engines with citations, 10 prompts included, extra prompts priced by the task.