Before I ever speak with a team, my agent stack watches their market and builds them a live brief: competitor moves, AI-answer gaps, regulatory clocks, and broken funnels — every catch dated, sourced, and checkable in under a minute. This room is the body of work. Client-commissioned briefs stay private; what you see below is the public series plus the shape of what runs behind the curtain.
Buyers now start with an AI answer, and in most B2B categories nobody knows who actually owns it. Every Monday I map one: who the engines cite, which surfaces feed them, and what it means for the vendors who aren't in the answer.
Health systems signed 15+ enterprise AI deals in the first half of 2026 and stopped running pilots on the way. The AI answer those deals get scoped against names four vendors, and some of the strongest companies in the category aren't in it. The anatomy of a $100M+ answer.
Read the brief →One B2B category per week, chosen from live market motion: a funding wave, a buying-behavior shift, or an answer that just hardened. Suggest one — especially if it's yours.
Suggest a category →Each of these was built unsolicited for a specific leadership team and delivered as a live page. They stay unlisted — a brief built for you is yours. The shapes below show what a morning of surveillance catches.
Six dated signals: a $125M-backed rival signed an enterprise deal inside the account where this company had 22 agents live, the category's buying motion went enterprise-first with no pilots, and the AI answer named four competitors — not them, despite the best logos in the space.
Their three competitors raised ~$160M in five months, all aimed at a different buyer — and one acquired a company the day before this startup's funding news crossed the wire. The lane they actually own sat uncontested, and unclaimed. The brief re-runs weekly and has logged new catches since.
A company serving 40% of the ten largest customers in its market was absent or bottom-ranked in the roundup content AI engines cite — while a direct rival swept the category's AI awards. Plus one catch worth the whole brief: their own careers funnel was returning a blank page.
A diversified portfolio audited in one morning: the holding company's contact page had no form, no email, and no phone; a crown-jewel subsidiary's indexed About page 404'd while a staging clone competed with the real site; and a 185K-follower consumer brand had been dark for six months.
Regulatory enforcement dates, network monitoring thresholds, and quantified buyer pain mapped into a trigger-to-play table: which public event lights up which member of the buying committee, and the program that earns the meeting. The JD called it strategy; the brief shipped it.
The first public body-of-work page: partnership map, GTM operating console, a full deal anatomy with champion mapping. "A founding Head of Growth does not pitch a partnership with a deck."
See the build →Method, on every brief: public sources only, no insider data. Every entry dated and tied to the surface where it was found, labeled observable vs directional. Nothing invented, nothing embellished — the entire offer is falsifiable specificity. If a brief's subject fixes something it flagged, that's the point.
Ask four AI models who they recommend in your category and see where you stand — the same instrumentation layer the briefs run on, in minutes.
Run the free checkYour market under surveillance for two weeks: the dated signal log, the AI-answer map, the conversion breaks, and the trigger map for your buyers — delivered as a live page that keeps running. If it doesn't catch anything worth a meeting, you'll know your surfaces are clean, which is worth knowing too.
Brief my market