Lead Generation
My internship manager offered me a job. I said no — so he became my client instead.
A one-person performance marketing business, and the acquisition stack I built to run it.
- Status
- Sept 2026 — ongoing~50% gross marginSolo
- Metrics
- 10–20leads/day~50%gross margin
- Stack
- Next.js · Google Ads · Meta Ads · GA4 · GTM · Claude + MCP
01 — How It Started
The offer I turned down is the reason this exists.
I spent six months at Walter Learning, a French EdTech doing €20M a year, running paid acquisition. I managed up to €220K a month across Google Ads, Meta and third-party lead platforms, cut cost per acquisition by 13%, grew daily lead volume by 39%, and lifted landing page conversion by 23% through systematic A/B testing.
Near the end they made me a return offer. I turned it down — I was going back to school and couldn't see myself staying. My manager's response was to propose something else: register your own company, and sell me the leads. So I did, and he became my first client.
02 — What I Built
I run the whole chain myself — the pages, the campaigns, the analysis on top.
I build my own landing pages, run my own campaigns, and sell the leads they generate to training organisations. There is no agency and no platform in between, which also means I carry the cost when a campaign doesn't work.
To follow it I built a dashboard: active landing pages, leads, views, conversion rate, cost per lead, and which variant is currently performing best.
The analysis is automated. At Walter Learning I had built a Claude plugin to help me read the ad accounts, and it saved me a lot of time; this one is a stronger version of it. A scheduled job goes over the campaigns and checks the things that usually drift — search terms picking up queries I don't want, keywords whose performance is dropping, CPCs moving. It reaches Google Ads through the MCP, so I can also just ask it questions instead of exporting reports.
Following the accounts that closely gets better results than leaving the bidding on full automatic, which is most of the reason the margin holds.
At the other end of the funnel is the page itself: a three-step qualification form that routes someone toward the right training and the right certified partner.
03 — Where It Stands
Started in September 2026, one client, running. Leads sell at €29 and cost €10 to €15 to acquire, so roughly half of what comes in is margin — which is the only number that decides whether this is a business or a hobby.
It's small and it's recent. What it gives me that the internship couldn't is full ownership of the loop: I choose the campaigns, I carry the cost of being wrong, and I keep what works. I'm looking for the second client now.
Get in Touch