Chapter two · The symptom the owner feels
Growth Shouldn't Feel Heavy
More work meaning more people is not a law of business. It is what happens when the number of hand-offs per job stays fixed while the number of jobs goes up.

Wilfred Greyling
Systems & Infrastructure

TL;DR
Four things most owners feel and few name correctly. Growth needs proportionally more people because coordination scales with volume. Everything routes back through you because you are the only one who can find out. Problems get discovered late, which costs far more than the chasing did. And the software bill rises every time you hire. None of these is a management failure, and none of them is fixed by working harder at the same shape.
Why does more work always seem to need more people?
Ask an owner what happens when revenue goes up thirty percent and the answer is usually a headcount number. It arrives quickly, without much working out, because it is not really a prediction. It is a memory of the last time.
The mechanism underneath is simple enough to check on the back of an envelope. In most businesses the number of hand-offs per job is roughly fixed. A job comes in and it passes between a certain number of people and systems before it is done and paid for. That number is a property of how the business is arranged, not of how busy it is. So double the jobs and you double the hand-offs, and each hand-off needs somebody to make it.
Coordination therefore rises in step with revenue. Not the work itself, which is what the customer pays for, but the moving of information between the people doing it. That is why growth feels like weight rather than freedom, and it is why the feeling does not go away when everybody gets better at their jobs. Nobody was doing it badly. There is just more of it.
It is the plainest thing in this whole series and the one most likely to be nodded at rather than acted on: the next thirty percent of revenue should not need thirty percent more people.
Why does everything still come back to you?
Most owners have tried to fix this and have treated it as a delegation problem. They have read the book, held the meeting, drawn the responsibility chart, and six weeks later everything routes back through them again.
It comes back because it is not a delegation problem. It is an information problem wearing a delegation problem’s clothes. When information is hard to get, the authority to decide migrates, quietly and without anybody choosing it, to whoever can actually get it. That is nearly always the owner, or one long-serving manager who has been there long enough to know which spreadsheet is the real one.
Nobody is hoarding anything and the team is not short of initiative. They cannot see enough to decide, so they ask the person who can, and that person becomes the bottleneck by default rather than by temperament.
Which means the fix is not another conversation about ownership. You are not the bottleneck because you will not let go. You are the bottleneck because you are the only one who can find out.
What does it cost when the answer arrives late?
Far more than the hour somebody spent chasing it, and this is the part that rarely gets counted at all.
Anybody who has worked in manufacturing knows the shape of it: the cost of a problem rises the further down the line it is found. A fault caught at the bench is cheap. The same fault caught at final assembly is expensive. Caught by the customer it is a different order of expensive again, and some of that cost is not money.
It generalises to every operational business. Slow information does not only waste the time it takes to gather. It delays discovery, and delayed discovery is where the money actually goes. The quote that was wrong, found after it was accepted. The job running over, noticed at invoicing. The client quietly unhappy for a month before anybody hears about it.
The version that lands in a room is one sentence long. Today you find out on Thursday. The decision was needed on Tuesday.
Why does the software bill go up every time you hire?
Because most of it is priced per person, which prices it against the exact thing you are trying to do.
Per-seat sounds fair, and for a while it is. It stops being fair the moment you notice that a business is not only its employees. Contractors, suppliers, the accountant, the lawyer and the customers themselves all have reasons to see something the business holds. Every one of them is another seat, and a seat for somebody who logs in twice a month costs what one costs for somebody who lives in it.
So a decision gets made that nobody actually makes. Access starts being rationed on cost rather than on whether somebody seeing a thing would help. The contractor works from an export that was accurate on Tuesday. The customer emails to ask rather than looking. Your own staff become the workaround, relaying by hand what the software could have shown.
It is a tax on growing, and it is charged at precisely the moment growing is working. We ran into it in our own business and it is one of the reasons we moved off rented software: who sees the work now gets decided on whether it helps.
So is this a technology problem or a people problem?
Neither, quite, which is why it survives so many attempts to fix it.
It is not a people problem, because the people are doing sensible things given what they can see. It is not really a technology problem either, because every business in this position already owns a great deal of technology. Most of them own more of it than they need.
It is a problem about what is connected to what. Every symptom in this chapter is a coordination cost: the hand-offs, the routing back through you, the late discovery, the seat you did not want to buy. They are one thing arriving in four disguises.
Which raises the question the next chapter is about. If the systems are not carrying information between themselves, something is.