A $7M-a-year growth consultancy being rebuilt as a one-person operation, where every job function becomes a playbook owned by an AI employee that reads from one shared context layer
~$7M/yr consultancy with millions in annual expenses (the human version) · the one-person agentic version has no revenue disclosed
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Serge Gatari runs acquisition.io, a growth consulting firm. His stated numbers: $15 million made over the last four years, on a roughly $7 million a year company carrying "millions and millions of dollars of expenses every single year." Later in the same video he refers to "my $50 million business," which does not reconcile with the earlier figures; we flag it and use the numbers he states first and repeatedly. His motivation is unusually honest for this genre: AI was a threat to him before it was an opportunity. "When AI hit the scene, I saw that it was actually a big risk to my business as it would allow one person to be able to deliver the same value as my $7 million a year company, where I have millions and millions of dollars of expenses every single year. If one person can deliver the same amount of value as we deliver with millions of dollars of expenses every year, they're going to run me out of business." So he asked the inverse question — how do I run the business that made me $15 million as a one-person company — and spent, by his account, fifteen months and a lot of investor calls and podcasts looking for the answer. The video is his attempt to explain why the technology has never been better while the people consuming content about it are still broke. His diagnosis: they are copying Anthropic's business model instead of building one of their own. Anthropic is at a $30 billion run rate by commoditising intelligence at the infrastructure layer; you do not get rich reselling a commodity. What he has actually shipped is a platform he calls Cook AI, and a claim that he has figured out how to convert the firm. What he has not shipped in this video is any revenue evidence for the converted version.
6 more sections — the full story, the playbook, the risks, and the numbers. Free account unlocks everything.
Create free accountEmail code only. No password.The quick-reference card — acquisition channels, replication playbook, and risk map — unlocks with a free account.
Create free accountEmail code only. No password.Paste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea: a full build spec and an SEO growth plan. Copy, paste, ship.
Go Pro — $10/moCancel anytime.Every claim traces back to a source video.
Source video links unlock with a free account. Full transcripts are Pro.
Create free accountEmail code only. No password.Data credibility: Founder-reported, with a live inconsistency and a clear commercial motive. Serge Gatari states that he runs acquisition.io, has 'made $15 million over the last 4 years', and describes it as a '$7 million a year company' carrying millions in annual expenses — then later in the same video refers to 'my $50 million business', which does not reconcile with either earlier figure. The transcript never resolves it; we report the figures he gives first and repeats. Nothing is shown on screen: no Stripe, no P&L, no client list, no contracts. Critically, all of the disclosed revenue belongs to the human-staffed agency, while the subject of the video — the one-person agentic conversion and the Cook AI platform built to run it — has no revenue, no customer count, no pricing and no client outcome disclosed anywhere. His claim that Anthropic 'surpassed OpenAI ChatGPT in hitting a $30 billion run rate' is stated without a source. The video ends in a free workshop and a platform trial, and he states the workshop teaches building 'six-figure-a-month AI growth firms in 3 to 6 months', so the incentives around every unverifiable claim run one direction. What we rate as genuinely valuable and independently assessable is the architecture argument — skills, integrations, playbooks with AI employees, and a single shared context layer instead of per-agent memory — which stands or falls on its own logic rather than on his numbers.