Not subscribed? Sign up to get it in your inbox every week.
A friend got on a call with a cold-call agency last month. He needed meetings with VPs of Procurement at global Fortune 500 companies (the kind of buyers who don't pick up unknown numbers and don't read cold emails).
The owner of the agency told him he could book 60 of those meetings a month.
My friend laughed. He'd done outbound before and knew the math: 60 meetings a month with that title was a fantasy number. He called BS and asked what the catch was.
There was no catch, just outcome-based pricing. The agency only got paid for meetings that actually showed up.
He signed. By the end of day one he had 5 meetings booked, by the end of day two another 7, and two weeks in his calendar held 44 meetings with VPs of Procurement at global companies.
He got a quarter of pipeline in two weeks.
The reason that story is possible is that the agency owner has codified his entire playbook into an AI system that handles what senior managers and trainers used to do: targeting, messaging, objection handling, call scoring, and ongoing coaching.
He still has SDRs on the phones, but they're hungry early-career people he hires for raw phone skill rather than years of experience. His system trains them, scores them, and coaches them up to senior-level output in weeks instead of quarters.
This is a structurally different kind of company than the agencies you've worked with before, and it changes the math on every functional hire you're considering right now.
SEO agencies got the same memo
In 2022, when I was running Copy.ai, we partnered with an SEO agency. The deal was $8-10k a month for eight articles, with a four-person team on their side: a strategist, a writer, an editor, and an account manager. We thought it was a fair deal at the time, and it mostly was for that era.
Last month, a friend signed with an AI-native SEO agency. He's paying $5k a month for hundreds of hyper-targeted articles, built to rank in both classic SEO and the newer AEO (making your brand show up in ChatGPT, Gemini, Claude, etc results).
The team on the agency side is the owner himself, working with a fine-tuned custom agent team he built using Claude Code.
That's the same budget bracket (actually a bit lower) for two orders of magnitude more output. The agency runs on a quarter of the headcount on their side, with none of the management overhead.
Past the AI sticker
For the last year or two, the prevailing advice has been to stop using agencies and bring everything in-house. AI tools made it cheap enough to build yourself, so why pay an agency markup?
That advice was written about traditional agencies, but AI-native agencies operate on different math. They run the same AI tools you'd use, with a senior operator who has codified their judgment, skill, and experience into the system. You get the leverage/speed of AI, the expertise of someone who's done it ten times before, and often a lower price.
The trap is to assume AI is doing only the obvious piece of the workflow ("AI writes the article" or "AI sends the email") and everything else looks like it did in 2022. The actual workflow has AI threaded through almost every step. The cold-call agency uses AI to build lists, train SDRs, run simulated buyer calls before any real one, and turn live transcripts into next-shift coaching. The SEO agency uses AI for keyword research, brief generation, article writing, fact-checking, and performance feedback into the next round.
The thing the old agency sold you was a team running a process. The thing the new agency sells you is the codified process running on autopilot, with a senior human in the loop for judgment calls and account ownership. The playbook used to live in the heads of senior people; now it lives in agents and workflows that the senior person built and tunes.
But how can you tell if an AI agency is actually AI native?
The simplest test is output volume. A real AI-native agency produces an order of magnitude more output than a traditional agency at the same price point.
Two cheap tests beyond raw volume:
Speed to first results. A traditional agency takes 30-60 days to onboard and another 60-90 to ship anything that matters. An AI-native shop has live work in your account within days.
Team size on the agency side. A traditional agency staffs your account with 3-5 people across functions. An AI-native shop usually has one senior person plus their system.
The clock beats the invoice
For the last decade, hire-vs-agency was a cost question.
The conventional answer was that it's cheaper to hire someone full-time at scale and to use agencies only for spike work or for expertise you couldn't recruit for. The math has moved, but not in the way most people are arguing about it: the cost gap closed, and the bigger shift now is time and speed.
Hiring an SDR: 6-9 months to fully ramp pipeline, on average. But that's assuming the rep works out, which most don't.
Signing with an outcome-priced cold-call agency: 44 meetings on the calendar in two weeks.
The question that actually matters now is "can I afford the time to hire?"
By the time your new hire actually starts, these new AI agencies have paid for themselves two or three times over and you've got a quarter of pipeline to show for it.
Want an intro to either agency?
The two agencies in this post are real, and every time I share their stories, I am asked for intros.
Fill out this 30 second form if you want me to make an intro.