I want to tell you the story I don't tell at conferences.
The version I usually tell goes like this: Realm scaled rapidly, raised $40M, built a product used in billions of devices, and was acquired by MongoDB. That narrative is accurate. It's also incomplete.
What the clean version omits is the chapter in the middle — after the Series B closed and before the acquisition process began — when we scaled the team to 68 people across San Francisco and Copenhagen, the fundraising market tightened, the C-round conversations that had been warm went cold almost overnight, and we didn't have the revenue to support 68 people without that capital.
We cut from 68 to 24 employees.
I was involved in all 44 conversations.
Here's what that week actually looked like.
Forty-four conversations over five days. Some were brief — engineers who had been with us for less than a year, who took the news with a professionalism that made it harder, not easier. Some were long — people who had moved from Copenhagen to San Francisco for this, who had turned down other offers, who had been building things they were proud of and had been given every signal that the trajectory was continuing upward.
The hardest conversations were with people I had hired specifically. I had made them a representation — implicitly, unavoidably — about what they were joining and where it was going. Some of that representation had turned out to be wrong. Not because I was dishonest, but because the fundraising market had changed in a way I hadn't prepared the company to survive.
One conversation I still think about was with an engineer we had recruited out of Copenhagen — someone I had personally persuaded to come to San Francisco, to take the risk, to join us at this stage. He had been with us less than eighteen months. When I told him, he didn't say much. He asked one question: "Was there something I could have done differently?" There wasn't. That was the point. He had done everything right. The company had made a structural mistake before he ever arrived.
The moment I remember most clearly is not the hardest conversation. It's the quiet after the last one — sitting in an empty conference room in the San Francisco office, which was going to need to be smaller now, and thinking: I did not protect these people well enough.
The diagnosis was simple. I had built a company that required the next round of capital to survive. When the market tightened, we had no survivable path at 68 people. The math was simple and it was brutal, and the only variable I could have controlled was burn rate.
I had known this was a risk in the abstract. Most founders do. The problem is that "default-alive" thinking is easy to practice when capital is scarce and hard to practice when capital is flowing. The round had closed, the market was hot, the obvious move was to scale.
The opportunity cost of not scaling — moving slower, being more conservative, having less capacity to build — was visible and immediate. The risk of building a company that couldn't survive without the next round was statistical and in the future.
I discounted the future too heavily.
What came out of that experience, permanently, is a principle I haven't compromised since: fundraising is optional, not a necessity.
This means something specific. It doesn't mean don't raise capital. We raised $40M at Realm. Capital accelerated things that needed to be accelerated. The question isn't whether to raise — it's whether the company is designed to survive without the next round.
The test is simple: if the next round doesn't close, can you reach cash flow positive before you run out of money? If the answer is no, you've built a company that can only continue to exist if external events cooperate. That's a fragile position, and the fragility shows up at the worst possible time — when the market tightens, which is exactly when the company most needs to be robust.
The 24 people who stayed were exceptional engineers and operators. The product was working. The technical foundation was right. All of that survived. What didn't need to happen was the 68-to-24 cut — if we had stayed leaner earlier, the question would never have been forced.
I've watched founders make the same calculation I made. Fresh capital, a compelling roadmap, the obvious move is to scale the team. The reasoning is the same: more people means more velocity means better metrics for the next round.
The specific error is treating the capital raise as a proof of durability. It isn't. A round closing means investors were confident in the trajectory at that moment. It says nothing about whether the market will cooperate eighteen months later when you need the next one.
The companies that survive the most brutal market corrections are almost never the ones that raised the most or scaled the fastest. They're the ones that maintained enough operational flexibility — enough of a path to profitability — that the market's cooperation was optional rather than required.
That's default-alive. Not a strategy you adopt because capital is scarce. A discipline you maintain even when capital is abundant.
Realm recovered. The team that remained built the product that reached 2 billion device installations. The acquisition by MongoDB validated everything the company had built.
But I carry those 44 conversations with me. Not as guilt — the company made rational decisions with the information it had, and the market changed in ways that were genuinely hard to predict. I carry them as a calibration.
Every time I look at a burn rate and think about whether to add more capacity, I remember what it feels like to be on the other side of a conversation where you're explaining to someone that the company can no longer afford them. The number on the burn rate spreadsheet is not abstract. It's a set of decisions that will affect specific people's lives.
Fundraising is optional. The people you're accountable for are not.