Jonathan Courtney went on Greg Isenberg's podcast and mentioned, almost in passing, that the campaign he was running that week should make $450,000.
The money comes from a webinar funnel he builds with Claude, using a four-step system he calls the promoter blueprint: traffic, holding pattern, selling event, conversion.
He's been running that same loop for 14 years at AJ&Smart, his Berlin agency. The only new part is that AI now does the production work in hours instead of days.
His opening shot at founders stung, though. He thinks most of us are using AI tools to hide from the actual job.
Here's the full episode, and below it, every number I could pull out of the 44 minutes.
The Agency That Doubled Revenue With a Third of the Team
Jonathan Courtney (JC to most people, @Jicecream on X) is the CEO of AJ&Smart, the Berlin studio that runs design sprints for LEGO, eBay, Lufthansa and N26. He also runs facilitator.com. He's told Starter Story the agency went from 36 people making 3 million euros a year to 14 people making 6.5 million, with about 4 million of that as profit in their best year.
So when he says "some of my businesses are multi seven figures," the receipts exist.
And here's the part that got me. He told Greg he uses AI tools "like 20% of how much you and your audience use these tools."
The guy making the most money off Claude in the room is the one spending the least time playing with it. His hours go into selling. The tools just compress everything around the selling.
Your Automations Might Be Procrastination Machines
JC's core argument: a CEO's real job is promotion, and AI tools "can be procrastination machines" for builders who'd rather optimize systems than tell anyone their product exists.
His analogy is a restaurant. Build the most automated, perfectly efficient restaurant on earth, tell nobody about it, and it dies in months. The automation never mattered.
Then he pointed at the people founders actually admire. Dario Amodei, Sam Altman, Pieter Levels: in his estimate, "at least 50% of their job is promoting their businesses." Levels' pinned post was a long-form promo for PhotoAI, and he published the sales bump it produced afterwards.
If you've spent three weekends wiring up agents and your revenue hasn't moved, that's the tell. I wrote about the same shift from the content side in founder-led content: the founders who show up everywhere are the ones people buy from.
JC frames promotion as oxygen for the business. Cut it off and the product quality stops mattering.
The Promoter Blueprint, Step by Step
JC normally sketches this on paper for clients. For the episode he vibe-coded it into a web page in about 20 minutes and shipped it to Vercel. Four steps:
1. Traffic. Only two flavors exist: organic (podcasts, events, social posts) or paid (Meta, TikTok, YouTube ads). Stop overcomplicating this box.
2. Holding pattern. Never push cold traffic straight at your product. Park people somewhere you can warm them up: a newsletter, a podcast, a YouTube channel, an X account. His biggest holding pattern is his email list.
3. Selling event. The bridge almost nobody builds. A live webinar or demo (he'd just been hit by a Superhuman Teams webinar campaign himself), an email sequence of three or four emails, retargeting ads built from a few thousand uploaded emails as a look-alike audience, or direct outreach. His favorite outreach move: scan the last 30 newsletter signups for an enterprise domain. One microsoft.com address = one booked enterprise call.
4. Conversion. A few buy. Around 90% of the ones who don't fall back into the holding pattern, where they wait for the next selling event. The loop never ends.
What makes this a system instead of a vibe is that he knows his ratios cold: "when I see that an extra 5,000 people have joined the newsletter, that means another 200 people will turn up to the next webinar, and that means another 12 people will buy something."
5,000 subscribers → 200 attendees → 12 buyers. At a high-ticket price point, that math funds an agency.
Most founders treat the holding pattern as a nice-to-have. It's the whole machine. If you want a deeper look at how to structure that warm-up layer, this breakdown of lead nurturing strategies covers the same ground JC draws by hand.
Inside the $450K Webinar Campaign
Now the headline number, with the honesty the headline leaves out.
At recording time the campaign had 4,000 webinar sign-ups. JC's words: "a campaign that will, I hope, make $450,000 this week... I can't guarantee... I think it will though. At least 350." Greg pegged JC's webinar campaigns at "between 250 and 500,000 dollars" and JC confirmed: "it's usually in that range."
So no, not a settled bank statement. A forecast from a guy who has run this exact play enough times to know his floor.
The interesting part is where the campaign got built: inside a Claude project he calls the "AJ&Smart marketing expert brain." It's loaded with his own books, his custom instructions, and a 17,766-line file of competitor marketing emails he's collected as a swipe file.
Live on camera, he asked it for lead magnet ideas tied to the episode. Seconds later: a promoter blueprint one-pager PDF, a "50 prompts for the promoter CEO" swipe file, and something called "the cave dweller audit: are you building or promoting?"
I'd download the third one, honestly.
The time compression is the real story. A full lead magnet campaign (idea, landing page, emails) used to take his team two to three days, sometimes a week once a designer and ClickFunnels-or-Framer decisions got involved. Now: half a day.
The One-Hour Claude Code Workflow
JC also showed exactly how he prepped for this episode, and it's the cleanest Claude-to-Claude-Code handoff I've seen.
Step 1: on his phone, a 15-minute voice brain dump, dictated through Wispr Flow, into a Claude project set up as his marketing assistant. Claude turned it into a scannable outline (he asked for "ADHD-friendly," which I'm stealing).
Step 2: at the end of every chat he asks Claude for "a pasteable set of instructions" to add to the project instructions. The project teaches itself. A separate chat built a reusable research file on Greg so that work never gets redone.
Step 3: the handoff. He asked Claude to "create a CLAUDE.md file that gives Claude Code the entire context of this project," dropped it in a folder, and told Claude Code to read it. Then he used plan mode and the AskUserQuestion tool to make Claude grill him on the idea before building.
Step 4: out came the Promoter Blueprint page (with a typo, he admits), pushed to GitHub, deployed on Vercel. Total time: about an hour, most of it running in the background while he did other things.
That grilling trick came from a Pip Decks co-founder, by the way. Same technique I covered in Claude Code as an AI employee: the tool gets dramatically better when you make it interview you first.
Abundance Beats Efficiency
JC kept hammering one strategic point: keep your team and use AI to multiply its output, instead of cutting heads to save money.
AJ&Smart kept their people and now effectively has "triple the size of the small team." They run five campaigns a month where they used to run one. Before, their Facebook ads got refreshed every six weeks because nobody had the bandwidth.
Two warnings came with it.
First, from his co-leader Laura: she spent three days building a custom proposal builder, then Claude itself pointed her to an off-the-shelf tool she could have used on day one. Always ask "does this already exist?" before you burn a weekend vibe-coding it.
Second: don't set up a thousand skills and sub-agents before you've shipped anything. JC was using AI as a search engine as recently as eight months ago. He learned chunked workflows by slamming into the context window on a real campaign, not by watching setup tutorials.
Start on a live project. The tools are forgiving enough to learn in production.
FAQ
What is the promoter blueprint?
The promoter blueprint is Jonathan Courtney's four-step revenue system: traffic (organic or paid), a holding pattern (newsletter, podcast, social) that warms people up, a selling event (webinar, email sequence, retargeting or outreach), and conversion. Non-buyers loop back into the holding pattern for the next campaign. He published the visual version at the link in the episode description.
Did the campaign actually make $450,000?
At recording it was a forecast, not a result. The webinar had 4,000 sign-ups, JC expected $450K that week with "at least 350" as his floor, and both he and Greg noted his campaigns usually land between $250K and $500K. Treat it as a realistic range from a repeatable system, not a one-off screenshot.
What is a holding pattern in marketing?
It's the warm-up layer between traffic and the sale. Instead of sending cold visitors to a checkout page, you park them in a newsletter, podcast or social following and nurture them until a selling event. JC's ratio: 5,000 new newsletter subscribers turn into roughly 200 webinar attendees and 12 buyers.
How much time should a founder spend on promotion?
JC's benchmark is at least half. He points at Dario Amodei, Sam Altman and Pieter Levels as proof that even the most technical CEOs spend 50% or more of their time promoting. If promotion isn't something you can stomach, his advice is blunt: find a co-founder who wants that job.
Steal the Playbook, Not Just the Tools
Every week on the Profitable Founder Podcast I sit down with bootstrapped founders doing $100K to $10M a year and dig out the systems behind the screenshots: the funnels, the ratios, the campaigns that actually ran.
If JC's blueprint was useful, the episodes go deeper.
