Most companies don’t die from lack of demand. They die from getting it.
How to scale a startup is really one question: what happens when the thing that was working stops working? The cruelty of it is that nothing looks wrong from outside. Revenue is up. The team is bigger. The press is better.
Inside, four systems are failing on different timelines, and no two fail at the same moment.
Here’s what breaks, in order, and the warning sign for each.
1. The Founder’s Calendar Breaks First
It goes before anything else, and almost nobody notices, because it doesn’t feel like a system failing. It feels like being busy.
At five people you sit in every conversation and that’s an advantage. At fifteen you sit in every conversation and it’s a bottleneck. At thirty you are the reason decisions take four days, because everything routes through the one person holding context on everything.
Warning sign: your team stops bringing you problems and starts bringing you approvals. They’ve learned that deciding without you carries risk. The moment that happens, your personal throughput caps your company’s speed.
The fix isn’t delegation. It’s context transfer. Delegating a task moves work. Transferring context moves judgment, so your team makes the next twenty decisions without you.
Rule: If you’re the fastest way to get an answer, you’re the slowest part of the company.
2. Hiring Breaks Second
Your first ten hires came through your network. People you knew, or people known by people you trusted. They arrived pre-vetted, culturally aligned, and willing to do whatever the week demanded.
That well runs dry somewhere between fifteen and twenty-five people. Then you hire strangers, and the hit rate collapses.
Founders respond by hiring faster, which is exactly wrong. A bad hire at forty people doesn’t just underperform. They hire more people like themselves, and now a whole branch of the org chart doesn’t work.
Warning sign: you describe a new hire as “we’ll see how they work out.” You never said that about hire number three.
What works:
- Write the scorecard before the job post. Five outcomes, ranked. Can’t name what success looks like in twelve months? You don’t know what you’re hiring for.
- Let someone who isn’t you make the final call on at least one hire per quarter. It’s the only way to learn whether your standard transferred.
- Fire in weeks, not quarters. The cost of a bad hire isn’t their salary. It’s the good people who leave because you tolerated them.
Rule: Slow to hire is not a virtue. Slow to decide is the mistake.
3. Cash Breaks Third, and Quietly
This one kills companies that were, by every other measure, succeeding.
Growth eats cash. You pay for inventory, headcount, and infrastructure before the revenue those things generate arrives. The faster you grow, the wider the gap — and profitable-on-paper businesses run out of money in the middle of their best year.
I’ve watched it happen to businesses with excellent margins and a full pipeline.
Warning sign: you check the bank balance more often than the P&L. That instinct is correct. Listen to it. Your P&L tells you a story about the past. Your cash position tells you about next month.
What to install:
- A rolling 13-week cash forecast. Weekly, not monthly. Thirteen weeks is long enough to see a problem coming and short enough to stay accurate.
- A hard cash floor. Pick the months of runway below which you will not go, and act at that line instead of past it.
- Your cash conversion cycle. Days from spending a dollar to collecting the revenue that dollar produced. Shorten it by fifteen days and you’ve funded a hire.
Rule: Profit is an opinion. Cash is a fact.
4. Decision-Making Breaks Last, and Worst
The endgame failure. This is the one that turns a fast company into a permanently slow one.
Early on, decisions happen in hallways. Somebody asks, somebody answers, done. Nobody writes anything down and it works fine, because everyone shares the same context.
Past fifty people that shared context is gone. The hallway conversation now excludes six people who needed to be in it.
So meetings appear. Then meetings to prepare for meetings. Then a process to manage the meetings.
Warning sign: two groups make the same decision and reach different answers. That’s not a communication problem. That’s a missing decision structure.
What to install:
- A single owner for every decision. Not a committee. One name.
- A split between reversible and irreversible. Make reversible decisions fast and alone. Give irreversible ones a week and a room. Most companies do this exactly backwards.
- The “why” written down, not just the “what.” A decision with no documented reason gets relitigated every six months by whoever wasn’t in the room.
Rule: Speed doesn’t come from working faster. It comes from deciding once.
How to Scale a Startup: Replace Proximity With Systems
Every one of those four failures has the same root cause. Something that worked because of proximity stops working when proximity disappears.
So scaling is one job: replacing proximity with systems, on purpose, before you’re forced to. That’s the whole argument of the Scale chapter in Start. Scale. Exit. Repeat., and it’s the stage where founders resist hardest, because building systems feels like bureaucracy when you’re used to speed.
It isn’t. Bureaucracy is what you get when you build the systems late, in a panic, after something already broke.
Take the four above and ask which one is currently your constraint. Not which is most broken — which is holding back everything else.
Fix that one. Then look again, because it will be a different one.
Scaling isn’t a phase you complete. It’s a bottleneck you keep moving. Just make sure you’re the one moving it, and not the last person to notice it moved.