Around 2013, Healarium was a small startup building an employee health improvement platform. Gamified goal-tracking for employees, insurance-side dashboards on the other end, a small R&D team.

Then we landed a Fortune 500 enterprise as a customer. That was the deal that made the company, and eventually the deal that ended it.

What Healarium Actually Was

The product sat inside the Obamacare wave. US employers were newly incentivized to help their workforce hit health goals, because doing so reduced their insurance costs. Healarium was the platform between the three parties. Employees tracked their goals, competed with each other in a gamified interface, and earned incentives (usually gift cards) from their employer for hitting targets. Insurance-side physicians got a dashboard that showed aggregate progress.

The engineering team was small. Django on the backend, Backbone on the frontend, a tight R&D group shipping features every couple of weeks. The kind of small team that can move fast because everyone knows every part of the system.

The Deal That Changed Everything

For a small B2B health startup in 2013, this was the biggest logo you could put on a slide. When they signed, everything changed. The tone of investor conversations changed. The pipeline slides changed. The team’s belief in the trajectory changed. This was the deal that proved the model.

It was also the deal where “having them as a customer” and “shipping the product they actually wanted” turned out to be two very different things.

What They Actually Wanted

The problem was not that they were hard to work with. The problem was that they were big.

Their list of requested changes was long, specific, and had a timeline attached to each item. They wanted the platform reshaped around their internal environment: their identity system, their compliance rules, their reporting formats, their approval workflows, their branding. Every part of the product had a “for them we would need this different” version.

Individually each item was reasonable. Collectively they were a full rewrite of the platform, pointed at an environment nobody else in our pipeline shared.

The Choice We Made

Our small engineering team faced a choice with no good version. We could ship their roadmap. We could ship the product the rest of the market seemed to want. We could not do both.

We chose them. It was the biggest deal on the books. It was the reference customer that would make every next sales conversation easier. It was the revenue story that made the company real to investors and to ourselves.

The alternative would have been to slow the engagement, push back on scope, and keep at least half the team building the general product. In hindsight that was the correct call, and in the moment it felt like leaving money on the table. Every founder in a small company that lands their first huge customer knows this feeling. It is not a rational calculation. It is a gravitational field.

What It Cost Us

Once the team was fully pointed at them, three things happened, in this order.

First, the general product stopped moving. Bugs the wider customer base cared about did not get fixed. Features the market was asking for did not get built. Every engineering hour was already allocated.

Second, the product started to bifurcate. There was Healarium-for-them, which had the custom identity flow and the custom reports, and there was Healarium-for-everyone-else, which had less and less of what it needed. Maintaining two forks with a small team is a specific kind of exhausting that does not show up in a status update.

Third, the team hit the ceiling of what they could sustain. Not a technical ceiling. A human one. The demands did not slow down, and there was no realistic path to hire fast enough to absorb them without eating the runway.

How It Ended

Eventually the demands outran what a small team could carry. Healarium as a going concern did not survive it. When the enterprise engagement wound down, there was no other product to fall back on, because the other product had not been touched in a year.

That is the outcome I still think about. It was not a technical failure. It was not a market failure in the usual sense. It was a strategic failure that looked, quarter by quarter, like the right call every single time.

The Big Customer Trap, 13 Years On

The Healarium story is a specific version of a pattern I now watch play out in almost every AI-first startup I talk to.

The shape is identical. Small team ships an early AI product. Some enterprise (a bank, a health system, a Fortune 500) runs a pilot, likes what they see, and offers a real contract. The contract comes with a list. On-prem deployment. A custom fine-tune. A private evaluation harness. Their brand, their guardrails, their identity system, their compliance framework, their reporting shape. Their timeline.

Every item on the list is defensible. The team says yes because the alternative is watching the deal die. The whole roadmap tilts toward one customer’s environment for the next six to twelve months.

What breaks is the same thing that broke at Healarium. The general product stops moving. The market signal from the other twenty conversations in the pipeline goes uninvestigated. The team hits the human ceiling of what a small group can carry. And when the enterprise pilot ends without a renewal (or the champion moves on and the deal quietly dies), there is no product left to sell to the market you stopped building for.

The lesson is not “do not sign enterprise deals.” Enterprise deals are how a lot of great B2B companies pay for themselves. The lesson is: when a small team lands one, most of the strategic work is defending the general product from the deal, not shipping the deal. That defense feels like leaving money on the table. It is the price of surviving the deal.

Two Things I Would Do Differently

With the benefit of thirteen years, here is what I would tell the version of us that had just signed that contract:

  • Split the team on day one, not day ninety. One group on the enterprise. One group on the general product. Not because you have the headcount, but because the alternative is that the general product silently dies while you are busy shipping features.
  • Negotiate scope, not price. Enterprise customers are used to being told no on features and yes on price. Small startups usually try the opposite. Every “we can build that” is a mortgage on the rest of the roadmap. Say yes to the deal, no to the shape, until you have to.

 

Let’s Talk

If your team is in the middle of the “should we take on this huge customer” conversation, or you already have and the roadmap is starting to feel like it belongs to them, that is exactly the shape of decision I work through with founders. Happy to walk through it with your specific numbers and stage, no pitch, just a working session. Reach out.