AI Visibility Monitoring
Your next client is asking ChatGPT, Perplexity and Google’s AI answers who to hire. We measure how often your firm is named, in what position, next to whom and how it is described. Then we repeat the measurement on a schedule, so you can see whether the work you do moves the number.
Why Measure It
The Shortlist Is Being Written for You
For legal, financial and advisory work, buyers ask an assistant for the best firms before they visit a website. The answer works as a referral, and it names only a handful of firms. It has happened to us: prospects have told us they checked several assistants before contacting Agile, and we were recommended in each.
It Is Invisible in Your Reporting
Search Console, GA4 and rank trackers do not show AI answers. A firm can rank well on Google and still be absent from what the assistants say.
One Answer Proves Nothing
AI answers change from one asking to the next. A screenshot is an anecdote. Decisions need a measured rate: the same questions, asked many times, from your market.
What We Measure
Discovery Questions
Questions your prospects ask without naming any firm, written the way people type them and grouped by service line. They show whether you are brought up at all.
Brand Questions
Questions that name your firm: what clients say about you, and how you compare with named competitors. They show how the assistants describe you.
From Your Market
Every question is asked as a user near your office, in your city, across your region or in your country’s largest cities, depending on how far you serve. Each one is asked several times on every assistant.
What You Receive
Discovery Visibility
The share of answers that name your firm for the service lines measured, in your market, with the margin of error stated.
Position and Share of Voice
Where you sit among all the firms named, how often you are named first, and which competitors are named most.
Visibility by Assistant
ChatGPT, Perplexity, Google AI Mode and AI Overviews measured separately, with Gemini and Grok added when your market uses them. They differ more than most people expect.
The Sources AI Relies On
Which directories, review platforms, publishers and competitor pages the assistants cite, and the sites cited next to your competitors but not next to you.
How You Are Described
The tone the assistants use about your firm, the strengths and concerns they repeat, with the actual quotes, and how you come out head to head with named competitors.
A Plan and a Trend
A prioritised action plan built from the evidence, then the same questions repeated weekly, fortnightly or monthly, so the report shows whether a change is real.
How an Engagement Runs
Scope
We read your website and agree the service lines, the competitors to watch and the market.
Question Set
We draft the questions and send them to you as a workbook. You approve, edit or reject each one, and only approved questions are measured.
Baseline
The first full measurement, typically around a thousand answers, establishes where you stand today.
Schedule
Measurements repeat weekly, every two weeks or monthly, always with the same questions, so the numbers compare like with like.
Review and Act
Each report comes with the findings and the recommended next steps. Your team or ours carries them out, and the next measurement shows the effect.
How to Read the Numbers
We would rather you trust a smaller number than doubt a bigger one. Figures cover only the service lines measured. Being named is not the same as being recommended. The report measures how often you are named and where you sit among the firms named. Where we show a range it is the 95% interval, and we only call a change real when it clears it. A 0% baseline is a starting point, not a verdict.
No one controls what an assistant says, so we do not guarantee placement. What can be influenced is the evidence the assistants draw on: your site, your reviews, and the directories and publications they cite. The report shows which of those matter in your market.
AI Visibility Monitoring FAQs
How is AI visibility monitoring different from tracking Google rankings?
Rankings show where your pages appear in a list. AI answers replace the list with a written shortlist of firms. You can rank well on Google and still be left out of that shortlist, or the reverse, so the two need measuring separately.
Which AI assistants do you measure?
ChatGPT and Perplexity through their official APIs with live web search on, and Google AI Mode and AI Overviews from Google’s live results. Gemini and Grok can be added when your market uses them. Each question is asked as a user in your market: ChatGPT and Perplexity receive your location through their official settings, and Google receives the coordinates of the point asked from.
Why ask the same question more than once?
Because AI answers change from one asking to the next. Each question is asked several times on every assistant in every measurement, five by default, so the report shows a rate with a margin of error rather than a single answer that may not repeat.
Can we see and approve the questions?
Yes. We draft the question set and send it to you as a workbook. Only the questions you approve are measured, and every report lists them with their results.
How quickly will we see change?
Changes to directories, reviews and published content take weeks to be picked up, so monthly measurement is usually the right rhythm to judge the effect. Weekly measurement suits firms running a campaign or watching a competitor closely.
How does this relate to your SEO and AI visibility audit?
Our Search and AI Visibility Audit includes a snapshot of what ChatGPT, Perplexity and Gemini say about your firm, inside a six-dimension review of your search visibility. Monitoring is the measured version of that snapshot: many more questions, asked repeatedly from your market and tracked over time.
Do we have to change our website?
Not necessarily. Often the quickest gains are elsewhere: the directories, review platforms and publications the assistants already cite next to your competitors. The report shows which ones matter in your market.
Can our partners or board see the report?
Yes. The report is a private, read-only link, optionally with a passcode and an expiry date, and a print-ready PDF. It is written for a non-technical reader and carries no internal costs or tool settings.
What does AI visibility monitoring cost?
There is no fixed package, because every programme is set up around your firm. The fee depends on the service lines and pain points we cover, how many questions we ask, how many times each one is repeated, which assistants and markets we measure, and how often the measurement runs. Tell us what you want to track and we will send you a quote before anything starts.
AI Visibility Insights
from Our Blog
Award-Winning Search Agency
Winner of the 2023 Clutch Global Award for SEO. See our press and recognition. Our services are recognised for excellence in both traditional SEO and AI search optimisation, and we were an early adopter of GEO for professional services.