AI Search Intelligence
Somebody just asked ChatGPT if they should trust you. What did it say?
A one-time Pilot Baseline Report shows you, in your own category, against your real competitors, exactly what ChatGPT, Claude, Perplexity, and Gemini are already saying, checked enough times to trust. $497, runs spread across 10-15 days so it isn't just one volatile day's reading, disclosed methodology included, no subscription required to see it.
11,800+ AI answers checked across every Pilot and Baseline report run so far. About 200 of those are yours, in a Pilot.
Built on the same track record that took Atlan's organic pipeline 7× from Series B to Series C: 10+ years running this exact playbook, not a first attempt at it.
What a Pilot actually finds
Real findings from Pilot and Baseline reports already run, rounded and de-identified to protect client confidentiality. The format is exactly what your own Pilot ships.
Enterprise SEO / AI-visibility platform
"Named in the large majority of unbranded answers, and recommended in almost none of them. The gap wasn't awareness. It was trust."
Share-of-voice tracking platform
"Functionally invisible on the single highest-intent category prompt: the exact one a real buyer types right before picking a vendor."
Compliance automation / GRC platform
"Tied for the category lead on general awareness prompts, but mention rate roughly halved on the two prompts closest to an actual buying decision."
What this isn't
"Just another Profound/Ahrefs-style dashboard."
A dashboard shows you the wound. This tells you where to put the bandage. Every finding ships with a Stage-mapped action, not just a number.
"The same thing as Consensus Monitoring."
That's for clients already fixing their foundation. See Consensus Monitoring. This is for anyone who wants to know what the market's telling AI before they've fixed anything.
"A predictor of leads or revenue."
No one can honestly promise that yet, including us. We'll tell you what's true today and what's trending. Never a number we can't back up. Not ready to spend on a Pilot yet? Start free with the Access Check instead.
How a report actually gets built
Four steps, in order. The discipline is the product, not the dashboard on top of it.
01
Prompt sourcing & validation
Real buyer language (search console queries, sales-call phrasing, review-site language, category-anchor comparisons), validated with you before a single run happens, not guessed at.
02
Multi-run sampling, across engines
Dozens of runs per prompt, per engine, spread over a rolling window: 2-4 weeks for the full Baseline, 10-15 days even for a one-time Pilot, because a single day's reading is exactly the kind of noise this discipline exists to catch.
03
Synthesis & Stage-mapped actions
Every finding gets mapped to a specific stage in the AEO/GEO sequence (an entity gap, an extractability gap, an earned-media gap), with an owner and an effort estimate, not a generic bullet.
04
Report delivered, then alerts on Intelligence
A narrative report either way. On the ongoing Intelligence subscription, that's monthly, plus a real-time alert the moment a competitor gets newly recommended or you get quietly dropped, not folded into a dashboard on a timer.
The receipts
Why the methodology is built this way: cited, not asserted.
↗
<1%
chance two runs of the same AI prompt return the same brand list: why no finding here is based on a single check
SparkToro/Gumshoe.ai, 2,961 runs
↗
65%
day-to-day source turnover in AI answers: why every trend needs a multi-week window, not a weekly snapshot
University of St. Gallen, 45-day study
↗
$154.5M
raised by the best-funded vendor in this category: per independent review, the one with the least disclosed scoring methodology of the vendors compared
rkarthik.in, vendor methodology comparison
Two ways to start
A one-time Pilot to see if this is real, or the ongoing Intelligence subscription once it is. Priced separately from the AEO/GEO retainer work. This is market intelligence, not implementation, and doesn't require a prior engagement.
Pilot Baseline ReportStart here
$497one-time
6-7 prompts · ~200 runs · 10-15 days
The real thing, at a fraction of the size: enough to surface one true, specific finding about how AI actually talks about you today. Same disclosed methodology as the full Baseline: sample size, run count, per-engine breakdown, one prioritized action.
If it doesn't surface one thing about your AI-search presence you didn't already know, it's a full refund. If you upgrade to Intelligence within 30 days, the $497 is credited toward your first month.
Get your Pilot Intelligence
$2,000$2,750/mo
Ongoing · 24-32 prompts · thousands of runs/mo
Everything the Pilot shows, scaled up and run monthly: full competitive landscape, citation-source analysis, buyer-language tracking, and Stage-mapped actions, plus real-time alerts the moment a competitor gets newly recommended or you get quietly dropped. Not a score. A to-do list, every month.
If the first report doesn't surface one finding you didn't already know, that month's free. Founding-client pricing available for a limited number of new clients, in exchange for a public case study on completion.
Talk about Intelligence +$750/mo add-on available on Intelligence for logged-in Google AI Overviews coverage: the one surface with no public API, kept as a separate line item rather than silently folded into the standard number.
Ready to see one
Most conversations start with a Pilot.
Tell me your category and who you think you're competing against for AI-search attention. I'll tell you honestly whether a Pilot is the right first move before either of us commits to anything.
Questions I get
What exactly is in the $497 Pilot?
The same structure as every report: headline findings with the exact evidence behind them, a competitive-landscape snapshot naming who AI names ahead of you and by how much, the real buyer language behind the prompts tested, where AI is actually pulling its citations from, and one prioritized action with an effort estimate. Smaller sample than the full Baseline (6-7 prompts, roughly 200 runs, spread across 10-15 days so it isn't just one volatile day's reading), but the same disclosed methodology, not a stripped-down teaser.
If I upgrade to Intelligence, does the Pilot cost carry over?
Yes. The full $497 is credited toward your first month of Intelligence if you upgrade within 30 days of receiving your Pilot.
How is this different from Consensus Monitoring on the methodology page?
Consensus Monitoring is a narrow add-on for active Growth Retainer clients: it tracks your own mention frequency, consideration-set trend, and factual errors, and I won't sell it before your foundation is fixed. AI Search Intelligence is a broader, standalone product: buyer language, competitor movement, and category whitespace across the market you compete in, closer to market research than implementation monitoring, and it doesn't require a prior engagement.
Does this predict future leads, citations, or revenue?
No, and no vendor in this category can honestly claim that. No published study establishes that an AI-visibility score measured today predicts a business outcome tomorrow. Every report ships with sample sizes and run counts next to every claim, and a confidence note on what the data can and can't support. That's the standard, not a caveat buried in fine print.
Which engines do you actually track?
ChatGPT, Claude, Gemini/AI Mode, and Perplexity via documented APIs. Google AI Overviews has no public API: that gap is disclosed, not silently substituted with a different surface. Bing Copilot and Brave Search are covered where a reliable collection method exists, at lower confidence, and that's stated plainly in every report.
What if a single week's numbers look noisy?
They will, sometimes. Independent research found under a 1-in-100 chance two runs of an identical prompt return the same brand list. No trend gets reported off less than a 2-4 week rolling window, and every number ships with the run count behind it so you can judge the confidence yourself instead of taking a single percentage on faith.
Is the methodology actually disclosed, or is that just a claim?
Prompt count, runs per prompt, per-engine collection method, and known gaps are published for every report, the same bar this site has held other vendors to. If a number can't be traced to a disclosed sample, it doesn't go in the report.
AI Search Intelligence. Built on the same evidence discipline as everything else here, not hype.
A good place to start
Your competitor is the answer
every AI gives your buyers.
A $497 Shortlist Scan tells you for certain in three days: real buying-stage prompts, run cold, three times each across ChatGPT, Perplexity, Gemini, and Claude.