Scale-up

In every earlier stage the problems a company worked on were outside it: what to build, whether anyone would buy it, whether the same method kept selling. Find the answer and you moved to the next box; fail to find it and you stopped there.

But somewhere between forty-five people and a hundred and fifty, the reason companies fall over changes. Demand is plentiful and output stops growing with headcount. Until yesterday the founder was involved in every decision and that was the fastest arrangement; now a queue forms in front of that person. The same feature gets built by two teams, and someone who joined three months ago still doesn't know who to ask. None of that shows up as a missed target until a release is already late.

Scale-up is the name for that stretch. The work here isn't finding a new market but rebuilding a method that already works so it doesn't collapse as people are added. Which is why the task list in this stage holds more organizational items than product ones: who decides what, when a new hire starts carrying their own weight, and where the information is written down.

The condition is that the way to sell is already known, and the constraint has moved from demand to the ability to execute.

1. The scale steps up an order. Headcount reaches the hundreds and revenue grows two to three times a year. The money that supports it usually arrives as a Series B or later, and a large share of it goes to people rather than to product or advertising.
2. The bottleneck moves to people and information. The number of people one manager holds is their span of control, and past five or six nobody can track everyone's work. So the first layer of managers appears in this stage, and growing those people internally or hiring them from outside becomes a task in its own right. Promoting the strongest individual contributor is the default move and the one most often regretted, because the two jobs share almost nothing.
3. What used to be handled in conversation starts demanding documents. At ten people you turned to the next desk; at a hundred the same decision happens between people who have never met. This is also the stage where tools like RACI — splitting who executes, who is accountable, who is consulted, and who is informed — start getting used. The purpose isn't to add process but to remove the time spent finding out who decides.
4. Hiring speed becomes the ceiling on growth speed. Hiring turns from somebody's side duty into a function, and the hours going into interviews and onboarding become a standing cost. Even at five hires a month, a new person spends about three months learning, so the execution capacity actually gained is closer to half the headcount added.

What comes after is either a large organization that runs on its own or an exit. Either way, four things get checked.

1. Growth stops being proportional to headcount. If revenue doubles when headcount doubles, revenue per person is flat, and that structure doesn't produce profit at greater size. Revenue per person starting to climb is the signal that the structure changed rather than just enlarged.
2. Plans start landing. Quarterly targets get set and actual results finish inside the range. Repeatedly hitting half the target or double it means the organization still can't be predicted.
3. The path to profit can be calculated. Running at a loss is fine, but the company has to be able to say in numbers which quarter it turns and under what conditions.
4. Internal controls and accounting get tidy. Aiming at a listing takes two to three years of preparation alone, and even aiming at an acquisition means contracts and revenue recognition have to survive verification.

Failing to clear these usually means slowing growth to fix the structure. Freezing hiring and reducing headcount are decisions that come up more in this stage than anywhere else, and they are easier to make early than late.

"What takes us down at this stage isn't demand, it's how fast we can hire."

The constraint is the organization, not the market.

"Let's stop settling this verbally and write down who decides."

There are too many people now for verbal agreement to hold.

"Headcount doubled and deploys per week didn't move. That's a structural problem."

Adding people isn't producing output, so the way of working needs examining.

As Pagedeck went from forty-five people to nearly a hundred, the first thing to break was product decisions. Prioritization that Mia had held alone split across three teams, and two of them started solving the same problem separately. One built notifications and the other built the same notifications on a different screen, and it surfaced two months later, just before release.

What they built afterward wasn't an org chart but a single table. It listed what each team owned, which decisions a team made on its own, and which ones the three teams made together. Grace and Mia spent two weeks on it, and the same feature was never built twice again.

The most common mistake is reading "grow quickly" as "hire quickly." The money is in and the demand is there, so adding people looks like growth itself. But as people are added, the number of connections needing coordination grows faster than the headcount. What one person has to know at ten people and at a hundred are different quantities. Hire without preparing to absorb them and a new person spends three months hunting for someone to ask while existing staff spend their time answering. So a hiring plan in this stage is set by the pace onboarding can absorb rather than by the number of open roles, and the limit is usually lower than the budget allows.

The other belief is that adding process slows things down. The memory of moving fast without rules makes documents and approval steps feel like a step backwards. But at a hundred people what actually consumes time isn't making decisions; it's finding out who makes them. With no designated place, the same item gets discussed three times and two teams build the same thing. Writing it down exists to remove waiting, not to impose control, and the organizations that genuinely move fast are the ones with few approval steps and no ambiguity about who decides.