Home/Cases/Case № 01
Case № 01 A regulated, underbuilt vertical DACH · 2025
Master use case · How I work

Signal to product-market fit.

From a cold market signal to 0% churn and paying users. Plus the operating model that made it possible.

Role · Product & Engineering lead Validation spend · < €10k Build time · ~2.5 months Outcome · 0% churn at PMF
The numbers — end to end Validation → PMF
01 — Validation cost
< €10k
Total spend to take the idea from signal to validated buyer.
02 — Ad spend
€0
No paid ads at all, just a cold-email campaign to a waitlist page.
03 — Build time
~2.5months
All in two-week cycles, after the operating model went in.
04 — Churn at PMF
0%
Paying users. Effectively perfect product-market fit.
One venture · one operator · one operating model Numbers, not a deck.
01 — The pitch ~45 seconds if you read it out loud
In one breath

Signal in. Paying users out.

Approx. 45s to read

I led product and engineering on one venture, from start to finish: from signal to PMF, under €10k, on 2.5 months of build.

It started with a market signal: money was flowing into a regulated market in DACH that almost no one had built software for. I turned that signal into a use case. The first audience I picked was the wrong one, and the interviews showed it, so I rebuilt everything around the real buyer one level up.

I cold-emailed a list I built by hand and sent it to a waitlist page. The result was about 60% open, 20% click, a real waitlist for €0 in ad spend. Then came the prototype, the decision-makers approved the pilot budget, and today there are paying users and zero churn.

The bigger story is what I had to change about how the team built to get there.

00 — Problem space

The stage I added.

Observe before anyone proposes a solution. Ground the rest.
01 — Idea

One-pager, one day.

Tagged by problem. Scored on signal. Gate vote.
02 — Concept

Hypotheses, measured.

Canvases filled. Every claim has a metric. Gate vote.
03 — Prototype

Ship the MVP.

Landing page · interviews · waitlist · live build. The pivot happened here.
04 — Pilot

Find PMF. Then scale.

Decision-makers approved · tranche-gated. Real paying users. 0% churn.
02 — The pivot The most important moment in the case

The pain was real. The budget was one layer up.

Stage 03 — Prototype
The workers have the pain.
The licensed operator one layer up has the power. And the budget.
What the landing page, the interviews and the cold emails told me, once I stopped reading my own guess into the data.

The workers on the ground had real pain: the language barrier, paperwork that wasn't their job, a clear problem a tool could solve.

But they had no buying power. They were low-paid workers in a price war, and the decision was never theirs to make. The pain was real, but the contract didn't sit with them.

One level up, the picture changed. The licensed solo operator runs their own book of clients, hands out the work, carries the legal risk and does the billing. That was the buyer all along.

The lesson

Pain ≠ buyer. Find the person with the power, the budget, and the consequence.

Wrong buyer
Worn work glove, a cheap pen and a plain envelope on cold linen — effort without authority.
Pain — realNot the buyer

The worker on the groundThe one with the pain

  • Real daily friction: language, paperwork, time
  • A low-paid worker in a price-driven market
  • The job isn't theirs to shape; it's handed to them
  • No budget, no say, no decision to make
  • The pain is real, but the contract isn't theirs
Real buyer
Fountain pen, keys, a leather notebook and a forest-green ledger in warm light — ownership and budget.
Budget — heldPays for it

The licensed solo operatorOne layer up · runs the business

  • Owns the plan, the schedule, the route, the day
  • Hands out the work but carries the risk
  • Runs a one-person business: billing, legal, logistics
  • Holds the budget and signs the contract
  • Knows the job, but not how to run a business. That's the gap
03 — How the team had to build differently The artifact that stayed

The pipeline is the what.
This is the how.

Old-school product and engineering throws specs over the wall, engineering disappears for six weeks, and what comes back is off. Three changes made the difference.

01 — Hypotheses, mapped

Sharper. Mapped. Not just named.

I coached the team to get specific, then map everything onto an Opportunity Solution Tree. It is a clear picture of problem, solution and experiment, so everyone could see which experiment tested which idea.

Method · OST + canvases · before anything is built
02 — Product · Eng as one

Reconnected to the customer.

The developers no longer get a feature spec. They get the customer context and the business goal, and they listen to the real interview recordings. They understand the problem, not just the ticket. We worked in Shape Up cycles: cutting scope is fine, adding time is not.

Outcome · No feature creep · laser focus
03 — AI as leverage

Boilerplate + supervised vibe-coding.

A boilerplate sits on top of the AI builders (Lovable, Cursor, Claude) with the important things already set up: security, keys, standards, checks. When it gets hard, the developers step in to clean up and rewrite, because the developers are the gate. They review and merge every change.

Stack · Lovable · Cursor · Claude · Supabase
Two stones bound by a single closed loop of forest-green thread on paper — one connected loop, not a relay.
CustomerIn the room
ProductFrames the problem
EngineeringShips the outcome
№ 02 One loop, not a relay. The interface is where work used to fall on the floor. Operating model · in practice
04 — The result Same team · same headcount · same funding

Different operating model. Different outcome.

We shipped v1 fast, pulled feedback straight from the waitlist, and had v2.0 out two weeks later, about four weeks in total. NPS moved up sharply, and then it held.

0%
Churn at product-market fit Nobody who paid churned. Users keep paying and keep using it. That is product-market fit, on about 2.5 months of development.
A single amber light trail rising from lower-left to upper-right on a dark ground — calm, compounding momentum.
App 01The day, ordered.

Plan and route every visit, so the day runs itself.

App 02Capture by voice.

Notes dictated between visits, with no German typing.

App 03The business, run.

Invoicing, legal records and the mileage log in one place.

If your room looks like this

The next case is yours.

It starts with one day inside your team, ends with a written diagnosis and a go / no-go decision, and either side can stop after day one.

Start a conversation → Read the Operating Model