What a growing company takes away is not authority but the speed of finding out
When a company has twenty people, the founder learns what has stalled from the corridor. A customer complains and she hears it the same day. A quote is priced wrongly and it comes up over dinner. At forty people the corridor lengthens, and at a hundred it is gone. The information still exists, but it reaches the founder in days, sometimes weeks. What she feels is usually phrased as "I am losing control", yet nothing has been taken from her authority. Her signature still counts and her decision is still the last word.
What has gone is the speed of finding out. By the time a decision reaches the founder's desk it may already have been taken, even carried out. It is experienced as a loss of authority because the outcome looks the same: the founder walks behind events in her own company instead of in front of them. The reaction is therefore predictable. We tighten. We want to see everything and we tie every decision to an approval.
Two sources of the same feeling
Telling them apart is simple. If the decisions the founder makes are being carried out, the problem is information, not authority. If they are being quietly ignored, that is a different matter and belongs to a different piece. In growing companies the first is far more common, but the fear of the second sits on top of it, and the founder carries two separate things as a single worry.
What tightening costs
The urge to see everything soothes in the short run and ties the company to the founder's calendar in the long run. The company moves as fast as the founder can approve and stops where she cannot. Control spread across everything is real nowhere, because one person's attention can truly read far fewer things than it can sign. The number of approvals rises, the quality of each falls, and the founder, believing control is in her hands, is only signing.
Control is not held everywhere; it is held at a few points chosen in advance
The method rests on a plain admission: in a growing company, what a founder can truly hold is limited, and the limit comes from the length of her day, not from her will. A leader who does not make this admission lets control go quietly and at random. A leader who does chooses what to release. The difference is the one between a disappointment and a design.
What a control point is
A control point is a decision or a signal, no more than three of them, that the founder does not allow to leave her hands. It might be a hard-to-reverse decision such as the treasury, the lowest acceptable standard for the product, or who gets appointed to the senior team. The choice differs by company but the test does not: these are the places where a wrong call costs the most and is hardest to correct.
The shortness of the list is not a constraint on the method. It is the method. A founder with five points can carry none of them, because each asks for attention every day. Three fit a weekly rhythm. In Is institutionalizing a loss of control we said that immediacy and control are different things. Here that distinction has a practical form: the founder who gives up immediacy keeps control only by holding it at a small number of points.
Three questions, three points
To choose them we ask the founder three questions:
- Which decision, if wrong, costs the most and is noticed the latest?
- Which standard is so tied to the company's name that it cannot be separated from the founder's?
- Whose arrival or departure changes the team's behaviour the most?
The answers usually settle into three headings, sometimes four. Removing the fourth is a small but hard step that decides whether the list works, because every heading once mattered a great deal, and dropping one means accepting that it will now sit in someone else's hands.
Control spread across everything is real nowhere.
For everything not chosen, the way information flows is designed separately
Everything outside the three points is released, and releasing has a rule: the form in which information reaches the founder, how often, and at what threshold are stated in advance. When this is not done the founder does not release, she only closes her eyes, and every day spent with closed eyes enlarges the worry of the next.
Summary, threshold, surprise
Three things are set up beforehand. The summary arrives weekly, on one page, written as a situation and not as numbers. The threshold says which situations skip the summary and go straight to the founder, for example an important customer's intent to leave or a warning sign from a key employee. The surprise is the assurance that news the founder does not want to hear travels the same route without penalty. The third is the most fragile, because once the founder reacts sharply to a bearer of bad news, the whole flow quietly shortens.
Information from the released ground arrives to the degree that the founder receives bad news calmly. It looks like a matter of tone, but it is information design: a sharp reaction makes the next summary polished and the next threshold high.
Every layer softens a little
In What growth erodes isn't direction, it's proximity we described how direction drifts from its source as a company grows. The same mechanism works on information: each layer softens what it passes on. The threshold and the no-penalty rule are the cheapest way to shorten that softening, because they need no new reporting system. They need one sentence, said once and then honoured.
The method is tested in two places: a monthly look back and the team's own mouth
Checking that the method works needs no new tool. Two tests are enough, and both fit into the founder's own calendar.
The monthly look back
At month end the founder lists the important decisions of the last thirty days and asks of each: when did I learn of this, from whom, and in what form. If one of the three points was learned about afterwards, it has in fact been released and should either be taken off the list or genuinely held. If a decision outside the points was learned about afterwards and cost nothing, the method is working. If it did cost something, the question is which link in the rule was missing: the summary, the threshold, or the no-penalty news.
The team knowing the three points
The second test is run through the team's mouth, not the founder's. We ask a manager from the team to name the founder's three points. If she cannot, the points have stayed private, and an unspoken control point is not a boundary for the team but a trap. When it is crossed, nobody knows what they crossed, and the founder produces an unfair outcome out of a fair anger.
If she can, the company's limit has become legible to everyone except the founder as well. In How to delegate we said that the limit of a handover has to be stated at the start. Control points are the founder's own share of the limits, and they work by the same rule: they help to the degree that they have been said aloud.
From the field
The subject usually enters the room as a complaint: "I no longer know what is going on." The complaint is accurate, and the remedy is rarely to collect more information.
What we do in the first session
In the first session we ask the founder to describe something she learned about afterwards, in the last three months, that bothered her. Then we ask: which of the three points was that outside of? Most of the time the answer is that the points were never chosen. The founder tried to hold everything, and so held nothing.
Not losing control while growing is not a matter of holding more; it is a matter of choosing what is held. How we work begins with that choice. The result often fits on the front of a card: three headings and a date to look at them again each month.