The founder bottleneck

The hub-and-spoke company

Draw the real org chart, not the one in the deck, the one traced from who actually asks whom before acting. In a bottlenecked company it's a wheel: the founder at the hub, everyone else a spoke. Marketing checks with the founder before launching. Sales checks before discounting. Delivery checks before pushing back on a client. The spokes rarely talk to each other, because the hub is faster, and because the hub has always been the place where the company's judgment lives.

The wheel has one property that matters more than all others: its maximum speed is the hub's calendar. Every growth lever the company pulls (more leads, more staff, more offers), adds spokes. None of them adds hub.

Why it happens to good founders

The bottleneck isn't a character flaw; it's yesterday's correct decision, compounding. Under $100K a month, routing everything through the founder is genuinely right: they have the taste, they carry the risk, and speed matters more than scalability. The company survives because of the hub.

The failure is silent because nothing announces the transition. No memo arrives saying the company now needs an operating layer. Revenue keeps growing, for a while, on sheer founder effort, which is precisely what disguises the ceiling as a rough quarter. Most founders discover the bottleneck the way you discover cold water: a failed senior hire, a missed holiday, a health scare, or the quiet realisation that they've become the least-leveraged person in their own business, doing $15-an-hour work between $10,000-an-hour decisions.

What it actually costs

Three costs, in ascending order of size and descending order of visibility:

• The latency tax, visible. Every decision queues behind the founder's calendar. A day's delay per decision, across hundreds of decisions a quarter, is weeks of company-wide idle time nobody logs anywhere.
• The parallelism ceiling, semi-visible. The company can only advance on as many fronts as the founder can hold in their head. Initiatives don't fail; they just never start.
• The key-person risk, invisible until it isn't. Pricing logic, client history, the real process: all resident in one skull. Every banker, buyer and insurer prices this risk even when the founder doesn't.

The two-week test

The cleanest diagnostic in the category: could you take two weeks off, unreachable, without the business losing money or clients? Not "would it be stressful": would it break. If the honest answer is no, the company doesn't have an operating system; it has your presence, and it's renting that presence at the price of your life.

Founders who pass the test aren't less involved. They're involved in different things, the decisions that actually need an owner's judgment, rather than the forty a week that only need someone's.

Dismantling it, five moves, in order

1. Turn recurring decisions into written policy. Every decision you've made three times is a policy waiting to be written. Discount limits, refund rules, hiring bars. Write them once, delegate them forever.
2. Build the one dashboard. A single place where the numbers that matter live, updated without you. Most founder-arbitration exists because nobody shares a source of truth to argue from.
3. Name an owner for every seam. Marketing-to-sales, sales-to-delivery: the handoffs are where disputes breed and where the hub gets pulled back in. Each seam gets one owner, in writing.
4. Install a weekly decision cadence. Forty-five minutes, run on the dashboard, decisions logged. The meeting exists so decisions have a place to happen that isn't your inbox.
5. Document as the default. If it isn't written, it didn't happen. SOPs aren't bureaucracy. They're the mechanism by which judgment leaves your head and survives contact with a new hire.

The self-diagnostic

Count how many of these are true:

1. My approval is needed for any spend over a trivial amount.
2. I am in every hiring decision, at every level.
3. Client escalations come to me by default.
4. Two departments have an unresolved dispute I'm avoiding.
5. I couldn't produce our three key numbers right now without asking someone.
6. At least one critical process exists only in my head.
7. My last real holiday involved my laptop.
8. The team's most common question is some form of "what do you want to do?"

Zero to two: you have an operating layer; protect it. Three to four: the wheel is forming: the five moves above, run yourself, will hold it. Five or more: you are the operating system, and the honest question is whether you'll dismantle it yourself in eighteen months of evenings or bring in someone who's done it before in ninety days. That second option is what this practice does, and either answer beats the status quo.