From 0 to $4M ARR with two engineers: how Layla built Greenmark

The fight that almost ended the company was about hiring. Our lead investor — a partner at the Munich fund — wanted us to deploy the capital fast. Two AEs. A head of marketing. A second backend engineer. A designer. Twelve hires he'd costed out himself in a spreadsheet he sent me at 11pm one Sunday.

I looked at the spreadsheet and felt physically ill. Every hire he'd listed would have come with a manager, a tool budget on top of our existing Notion, Slack, and GitHub stack, and a quarter of onboarding before they shipped anything. The math said our €280k MRR would not pay for them; the next round would. I'd watched two friends' companies die that exact way in 2022.

§Saying no to the spreadsheet

I told him no on a Tuesday call. He told me, calmly, that I was leaving "an order of magnitude on the table." Two weeks later he stopped replying to my updates. I found out from another founder that he was telling the climate-fund circuit I was "uncoachable."

Petar and I had our worst fight that month. He'd wanted the engineer hire. We didn't speak for a weekend. I later met the founder of Tarmac who hit a similar capital-deployment fork and his framing — that capital deployment is not progress — was the language I wished I'd had on that call.

The unlock came from a feature we almost didn't ship. In October 2024 a customer asked whether we could pull data directly from their fleet-card provider, DKV. I said no twice. Petar built it anyway over a weekend because he was annoyed.

The DKV connector cut onboarding for any company with a vehicle fleet from seven days to about ninety minutes. We added Shell Card, UTA, and Circle K within the next month. Logistics, construction, and last-mile delivery customers started signing at twice our previous rate. We went from €31k MRR to €110k MRR in the four months after that connector shipped.

We never made the twelve hires. Our team is still three people plus two contractors. As of April 2026 we're at €330k MRR — about $4.3M ARR — with a 4% logo churn rate annualized and a contribution margin north of 80%. We pay roughly €3,200/month total across AWS, Vercel, Stripe, and Resend for the entire infrastructure footprint. Petar and I still pair on every customer call that lasts longer than fifteen minutes.

The Munich partner emailed me last month asking whether we'd raise a Series A. I haven't answered yet.

§What we did with the runway we didn't burn

In the four months after the DKV connector shipped, our blended CAC fell from €4,800 to €1,650. The reason was unglamorous: every fleet customer who signed referred at least one peer. Mazars, PwC Belgium, and Forvis Mazars Germany started recommending us by name to their mid-market clients because the audit handoff went from a six-week back-and-forth to a one-call sign-off. Trial-to-paid conversion climbed from 22% to 41%. Logo churn dropped to 4% annualized.

We made exactly one hire: a part-time auditor-relations contractor named Anke who did nothing but build the relationship with the Big Four regional partners. That single role moved more pipeline than the four-AE team our investor had wanted. Our net revenue retention sat at 118% by April 2026, mostly from existing customers adding sites and modules. We turned off HubSpot that quarter and replaced it with a Postgres view plus a small Resend sequence — the CRM had been costing us €1,400 a month and three hours of weekly maintenance for no measurable lift. The team is still three full-timers and two contractors.

1. 1
   Capital deployment is not progress.

   Every hire is a multi-year liability disguised as growth. We grew faster after I refused to spend the round than we did the quarter before it landed.

2. 2
   Build for the regulatory deadline, not the regulatory framework.

   Customers don't wake up wanting CSRD compliance. They wake up wanting their auditor to stop emailing them. Aim at the second one.

3. 3
   The connector your customer asks for twice is the product.

   I said no to DKV twice. The third request — and Petar's stubbornness — quadrupled our growth rate.

4. 4
   "Uncoachable" is sometimes the right answer.

   Investors optimize for outcomes across a portfolio. Founders have to optimize for survival of one company. The advice is rarely calibrated to your specific risk.

5. 5
   A small team is a moat.

   Three people who know every customer, every line of code, and every invoice will out-ship a fifteen-person team for at least eighteen months. Don't trade that lightly.

What this looked like in practice.

The thing nobody tells you about *capital deployment* is that the cost of a hire is not the salary. It is the salary, the manager you eventually have to add, the tool budget on top, the recruiting spend, the equity grant, and the six months of context-loading before the person ships. We modeled it once and a single mid-level engineer in Berlin landed at €240k all-in for the first year. That number is what made saying no easy.

The thing nobody tells you about *building for the auditor* is that auditors talk to each other constantly. We never paid for ads. The Mazars network became our distribution. Aim at the person whose pain compounds across your buyer's vendor — they are a lever, not a customer.

The thing nobody tells you about *the connector your customer asks for twice* is that the second ask usually comes from a different customer, which means the demand is real and not idiosyncratic. We now triage every "can you connect to X?" request by counting *distinct accounts* asking, not total tickets.

The thing nobody tells you about being *uncoachable* is that the label gets walked back the moment your numbers work. The Munich partner's email last month was warm.

The thing nobody tells you about *small teams as moats* is that the moat is not speed — it is the lack of internal politics. Three people don't have a roadmap argument; they have a five-minute conversation.