The Problem
Eighteen months in, we had eighteen active pilots. Healthcare clinics. A boutique hedge fund. Two community banks. A K-12 charter network. A telehealth provider in Portugal. A crypto custody startup. We had decks customized for each vertical. We had a sales engineer flying every other week.
We had also closed exactly zero of them into paying contracts.
Each pilot had a unique blocker. SOC 2 evidence requirements were different from HIPAA's. The hedge fund needed SEC-style audit trails our schema couldn't model. The charter network had a procurement cycle that ate two quarters before anyone could sign a check.
We weren't building a compliance platform. We were running eighteen unfunded consulting projects.
The Journey
Ledgerline started as a tool I'd hacked together on weekends to automate evidence collection for SOC 2 — the one thing my previous company had spent six figures on every year. It worked. My old CTO became our first design partner.
We raised a $2.4M seed in six weeks. The deck said "horizontal compliance platform for the post-cloud era." It tested well with investors. It tested terribly with reality.
In our first year we hired two account executives. Their job was to "expand the wedge" beyond SOC 2 into adjacent frameworks. The AEs were good. They booked demos with anyone who breathed. Healthcare. EdTech. Web3. The pipeline graph looked beautiful.
But every demo led to a custom build request. Every pilot we won added a new line item to the engineering backlog. We kept saying yes because saying yes felt like progress.
The Struggles
The cofounder fight came in month sixteen. My cofounder, who ran sales, wanted to keep "going wide" — she argued our pipeline diversity was our moat. I'd just spent a Sunday tracing every customer interaction in our CRM and realized 71% of our engineering hours were going to features used by exactly one pilot.
We stopped speaking for four days. I went to bed at 2am most nights reading vertical SaaS post-mortems and found the same shape every time: companies that tried to serve every industry built mediocre products for all of them.
The board pushed back when we proposed narrowing. "You're shrinking the TAM," one of them said. I sent him a one-line email: "The TAM I have today is zero. Pick a non-zero number."
We agreed to a 90-day experiment. Pick one vertical. Sunset the rest. If revenue didn't move, we'd reconsider.
The Breakthrough
We chose Series B-stage SaaS companies preparing for their first SOC 2 audit. Boring. Specific. Painful enough that buyers had budget. Repeatable enough that one playbook would work across customers.
We rewrote the website in a single afternoon. We canceled fourteen of the eighteen pilots — keeping only the ones that fit the new ICP — and refunded their pilot fees. Half of those founders thanked us. Two became referenceable case studies later.
Within sixty days we'd closed seven new logos at an average ACV of $34,000. The sales cycle dropped from 92 days to 28. The engineering team — which had been triaging ten different roadmaps — collapsed into one. Velocity tripled, measured in PRs merged.
By the time the 90 days ended, we had $260k in newly-booked ARR. The board stopped asking about TAM.
The Lessons
- 1Pipeline diversity is not a moat — it's a tax.
Every additional ICP you serve fragments your roadmap, your messaging, and your hiring plan.
- 2A pilot is not a customer.
If you can't articulate the post-pilot contract motion in one sentence, you're running consulting projects with extra steps.
- 3Refund money to focus.
Sending fourteen pilots their money back was the most expensive thing we did. It was also the cheapest, because it bought us a year of clarity.
- 4Boring verticals close.
The most exciting industries usually have the longest sales cycles and the highest custom-build risk. Boring + budgeted + repeatable beats sexy + pioneering every time at our stage.
- 5The TAM argument is a status game.
Investors want big TAM stories. Customers don't care about your TAM. Pick your audience.
