Skip to content
FOR FOUNDERS ESTABLISHING AI VISIBILITY
STARTER
$24/month
- 1 brand · 10 tracked prompts
- ChatGPT + AI Overviews · daily
- 10 research searches / month
- Snippet + crawler tracking
01INTRO
WE ASK CHATGPT, PERPLEXITY, GEMINI AND AI OVERVIEWS WHAT YOUR BUYERS ASK, EVERY DAY, AND COUNT HOW OFTEN YOU COME BACK
02AREAS
Where you stand
Your mention rate across four engines, multi-run, with 95% confidence bands and the position you hold in the answer.
Who is beating you
The rivals AI recommends instead, the prompts where they win and you are absent, and the sources they get cited from.
What to fix next
A weekly roadmap scored on demand against the visibility you have, grounded in the domains cited for that topic.
Whether it worked
Movements that clear the confidence band, and every prompt bucketed winning, contested, losing, or too early to call.
Who AI sent you
The AI-referred visitors your analytics files under “Direct,” labeled certain or likely, never mixed together.
What agents read
GPTBot, ClaudeBot and PerplexityBot reads per URL, checked against published IP ranges, joined to what got cited.
03PROCESS
We turn noisy AI answers into a measurement you can act on.
- .01
Tell us your brand
Setup >
Paste your URL. We draft the questions your buyers ask AI and you approve the ones worth tracking. Five minutes.
- .02
We interview the engines
Measure >
Every day we put your questions to ChatGPT, Perplexity, Gemini and AI Overviews, several runs each, because answers change.
- .03
Add the snippet
Detect >
One line of code for AI-referred visitors, plus a drop-in server-side helper so agent reads get counted, not guessed.
- .04
Read Monday’s briefing
Act >
What moved, who overtook you, which pages got cited, and three things to do about it. In a week with nothing measured we send nothing.
04RESULTS
1%
RUN-TO-RUN SWING
Ask an engine the same question ten times and one brand can appear anywhere from 20% to 80% of the time. We sample; we don’t guess.
1
ENGINES MEASURED DAILY
ChatGPT, Perplexity, Gemini, and Google AI Overviews: every tracked prompt, multiple runs, every day.
1%
CONFIDENCE BANDS
We report rolling mention rates with 95% confidence bands, and say “no significant change” when they overlap.
06AUDIENCE
Whatever you sell, AI is already recommending someone in your category
SaaS & apps
- “Best tool for X” answers
- Competitor comparisons
- “Alternatives to” pages
- Landing pages AI cites
Agencies
- Client workspaces
- White-label digests
- Pooled credits
- Movement they can show
E-commerce
- Category recommendations
- AI shopping referrals
- Agent crawl coverage
- Category citations
Local & services
- Local recommendation answers
- Directory citations
- Who AI recommends nearby
- Review-source tracking
Publishers
- Cited-source tracking
- AI crawler reads
- Referral traffic
- Topic coverage
B2B & consulting
- Vendor shortlists
- Category authority
- Named-in-answer rate
- Rivals on the shortlist
07STACK
One line of code for AI-referred visitors, plus a drop-in server-side helper so agent reads get counted, not guessed. Next.js, WordPress, Shopify, Webflow, Astro, Cloudflare.
08VALUE
.01
RATES, NOT ONE-OFF SCORES
.02
VERIFIED CRAWLERS, NOT CLAIMED
.03
AGENT READS, VERIFIED BY IP
.04
EU-FIRST, NO RAW IPS
MUCH MORE
WE LEAD WITH OUR RECEIPTS, NOT A HYPE SCORE
09METHOD
Multi-run sampling
Every tracked prompt, every engine, every day
Ten runs of the same question can put one brand anywhere from 20% to 80%. We report the rate and the band around it, never the lucky run.
11HELP
Once, sure. But answers change run to run, day to day, engine to engine. One check is an anecdote. We turn it into a measurement: multi-run sampling, every day, across four engines, with history you can act on.
SEO got you ranked in a list of links. This measures whether you get named in the answer above it, which your SEO tools cannot see.
Anyone can type “GPTBot” into a User-Agent header, and plenty do. We verify every crawler hit against the IP ranges OpenAI, Anthropic, Perplexity, and Google publish for their bots. Verified hits count; impostors get their own bucket.
Built EU-first: EU hosting, no raw IP addresses ever stored, no cross-site tracking. The crawler-tracking side processes no personal data at all (no humans, no cookies, no identifiers), so it works even under the strictest consent setup.
Not yet. We measure what the engines say, who they send you, and what their crawlers read. Joining that to revenue is on the roadmap, and we would rather say so here than show you a number we cannot stand behind.
No. We count sessions and landing pages, and we label the ones AI sent you. What we add is the half your analytics cannot see: which visitors arrived from an answer engine, and which AI agents crawled which pages.
Then you found out today instead of next year, and your weekly digest tells you which sources the engines cite in your category, so you know exactly where to earn your way in.
