Right now someone in your market is asking ChatGPT for the best option in your category. It answers with three or four brand names and a reason for each. If yours is on the list, you are winning trust and traffic that never shows up in your analytics. If it isn't, the buyer walks away with a rival's name and you have no record it happened.
Watch it happen and it looks like this:
Buyer asks: “best CRM for a small marketing agency”
Popular choices include HubSpot for all-in-one marketing, Pipedrive for a simple sales pipeline, and Copper for agencies built around Google Workspace. Most reviewers point content-led agencies to HubSpot first, then suggest Pipedrive if you mainly want a clean deal flow.
Your brand never appears. The buyer leaves with three competitors in mind, and nothing in your analytics shows the conversation took place.
For years the job was simple: rank on Google, where a buyer scanning ten links would find you. Now the buyer reads one AI answer and acts on it. When someone asks AI about your category, does it name you?
You cannot answer that from Google Analytics or Search Console. Neither tool watches AI answers.
Your buyers already act on AI answers
Ask ChatGPT, Gemini, Claude or Perplexity to recommend a product and you get a short list of brands, often with reasons and source links. Buyers treat that list as a vetted shortlist. They open one or two of the names and ignore the rest, which hands the first look to whoever the engine mentioned.
That shortlist forms before anyone visits a website. By the time a buyer reaches your homepage, they have already read an answer that shaped who they trust, and it may not include you.
Why can't you see what AI says about your brand?
Because no analytics product watches answer text. Your analytics count visits. Search Console shows Google queries and clicks. Neither one tells you that Gemini recommended a competitor on Tuesday, or that ChatGPT described your product with a feature you retired last year.
That gap costs you buyers before you know they exist. The channel runs every day, and most brands have no window into it.
What does AI visibility tracking actually mean?
AI visibility tracking runs the real questions your buyers ask, sends them to each engine, and records the reply: whether you are named, how often, in what context, and which competitors appear beside you.
Traditional SEO tools do a different job.
What SEO tools show
- Where your pages rank in Google
- Keyword volume and difficulty
- Backlinks and site health
- Clicks from the results page
What AI answer tracking shows
- Whether ChatGPT and Gemini name you
- How often you appear, and for which questions
- Which rivals get recommended instead
- How AI describes your brand and what it cites
Both matter. They answer different questions, and only one of them tells you what a buyer hears when they ask AI.
How do you track brand mentions in Gemini?
You send a fixed set of buyer questions to Gemini on a schedule and record which brands each answer names, rather than asking once and reading the reply. A single check tells you very little, because Gemini's answers move between sessions, accounts and locations, so two colleagues can ask the same question and get two different shortlists.
Before you start, work out which Google surface you actually mean. Three of the seven engines Koalr tracks are Google products and they behave differently:
Gemini is also the engine least likely to point at a company's own homepage. Across the 295,746 classifiable citations in Koalr's dataset, 5.6% of Gemini's citations resolved to a homepage, against 34.1% of ChatGPT's.
That six-fold gap has a practical consequence for tracking. A polished homepage is often enough to get named by ChatGPT, and it is rarely enough for Gemini, which assembles its answer from a wider spread of pages you mostly do not own. So when Gemini omits you, the fix usually sits off your own site, and you will only find it by looking at what Gemini cited instead. The full breakdown is in our study of 300,318 AI citations.
Tracking Gemini by hand means opening a fresh session for every question, recording the brands named, repeating it weekly, and doing the same again for the other six engines. That is the job Koalr's brand monitoring automates.
How Koalr tracks your brand across seven AI engines
Koalr checks ChatGPT, Claude, Gemini, Perplexity, Microsoft Copilot, Google AI Overviews and Google AI Mode. You give it a domain. It runs 25 real buyer questions across all seven and reports whether each answer names and cites you.
- 1
Enter your domain
No tag, no developer, no onboarding call. Paste your website and Koalr takes it from there.
- 2
Koalr asks the engines
It runs 25 real prompts from your category across all seven engines, the way your buyers ask them.
- 3
You get your score and gaps
An AI Visibility Score from 0 to 100, plus the questions where a competitor is named and you are not.
See where your brand shows up in AI search
Enter your website and get your AI Visibility Score across 7 engines. First score in under 5 minutes, no setup.
What you see: numbers you can act on
Koalr turns the raw answers into an AI Visibility Score from 0 to 100. It combines your mention rate, your citations, your share of voice against named rivals, and the tone of how each engine describes you.
Under the score you see the detail: which engines name you, for which questions, and which competitor keeps taking a mention you want. Add every brand you manage, so an agency watches a full client roster and each client's rivals from one place. You can report that score to a client or a board and track it month to month. For what a roster of client brands costs across the five main platforms, see the AI visibility tools for agencies comparison.
Setup takes minutes, not weeks
You skip the onboarding call and the developer. Paste a domain and Koalr returns a first score in under five minutes. The scan is free to run, and every plan starts with a 7-day free trial when you want scheduled tracking and history behind it.
Turn the data into more citations
Tracking is the start. The value is the list of questions where a rival is named and you are not. Each one is a page you can write or sharpen so an engine has a better source to quote.
Say Gemini keeps recommending a competitor for a question you should own. Open the answer, look at what that competitor published and where they earned their citations, then close the gap on your own site.
Koalr will not write that content for you. It tells you which questions to write for, the part most teams get wrong.
If you would rather start with a single manual check before setting up tracking, our AI visibility checker walkthrough covers how to test your brand by hand and what the result does and does not tell you.
Start watching your AI presence today
Most of your competitors are not tracking this yet, which is the opening. Run a free scan at koalr.ai and see where you stand across ChatGPT, Gemini and five other engines. Questions about your score go to hello@koalr.ai.
Frequently asked questions
Run the same set of buyer questions through Gemini on a schedule and record which brands each answer names, rather than checking once. Gemini's answers vary between sessions, accounts and locations, so a single check is a snapshot rather than a measurement. Koalr runs 25 real buyer prompts through Gemini alongside six other engines and reports whether you were named, who was named beside you, and which pages Gemini cited.
