Getting Out of the Owner Bottleneck: Delegation for Subcontractor Founders

There's a revenue band — for most trade contractors somewhere between $3M and $10M — where the company stops growing and nobody can say why. The market's there, the crews are good, the bond program has room. The constraint is quieter: every consequential decision still routes through one phone. Which bids to chase, what price goes out, whether the change order gets signed, who gets hired, whether the GC's revised schedule is acceptable — all of it waits on you, and you are fully booked.

The bottleneck isn't a character flaw. It's the residue of what made the company work at $1M: you were the best estimator, the best closer, and the best foreman, so everything routing through you WAS the quality system. But the same wiring at $6M means decisions queue for days, your best people learn that initiative gets overridden, and your own calendar becomes the company's critical path. The fix isn't working more hours — you've tried that. It's converting the judgment in your head into systems other people can run.

You've probably attempted delegation before, and it probably went like this: you handed someone a task, they did it differently than you would have, something went wrong or merely looked wrong, and you took it back — privately confirming that it's faster to do it yourself. That loop fails for a specific reason: you delegated the task without delegating the standard. The person never knew what a good outcome looked like, because the definition lived only in your head.

You cannot delegate outcomes you've never defined. Before any handoff works, the standard has to exist outside your skull: what makes a bid worth chasing (the go/no-go scorecard), what contract terms we accept and reject (the playbook), what a complete pay application looks like, how a handoff meeting is run. Written standards do two jobs at once: they let someone else decide the way you would, and they turn 'you did it wrong' — a personal fight — into 'we missed step four,' a process fix. The delegation problem is mostly a documentation problem wearing a trust costume.

The unit of delegation that actually works is the decision type with a threshold, not the task. 'Help me with change orders' delegates nothing — every instance still comes to you for blessing. 'You own change orders: price and submit anything under $10K yourself, bring me anything over, and never do added work without a signed directive' delegates something real: a category, a limit, and a rule for exceptions. The person can act without you, and you've bounded the cost of their learning curve.

Map your week for ten days and log every decision that touched you; then sort them into three piles. Only-you decisions: pricing strategy on major pursuits, hiring leadership, signing contracts, the banking and surety relationships. Threshold decisions: yours above a line, theirs below it — purchasing, change orders, schedule commitments, small-job bids. Not-actually-decisions: things you're reviewing out of habit that a standard could replace entirely. Most owners discover the third pile is the biggest — half the queue isn't judgment at all, it's approvals a written rule would handle. Kill that pile first; it's free capacity.

Here is the part nobody warns you about: delegation makes things slightly worse before it makes them better. The first quarter after a handoff, your PM prices a change order lower than you would have, the estimator chases one bid you'd have skipped, a schedule commitment gets made that squeezes a crew. This is the J-curve, and it is the tuition of building leaders — the same air-cover discipline that applies to new foremen applies doubly to new decision-owners. Correct privately, back publicly, and treat mistakes as calibration data for the standard, not evidence the person failed.

The fatal move is reclaiming the decision at the first error. Do that once and you've taught the whole company that authority is decorative and initiative is risky — you're back to being the bottleneck, plus now everyone knows trying to help you is pointless. The discipline instead: when a delegated decision goes sideways, fix the standard, not the org chart. Was the threshold wrong? Was the playbook silent on this case? Was the training thin? Each mistake that improves a written standard is a mistake that never repeats — which is more than can be said for the mistakes you make yourself when you're doing five jobs at once.

Delegation without a review rhythm decays into either abdication (you find out about problems at WIP-review time, months late) or reversion (people resume asking you everything because it's easier). The replacement for your open door is a short, fixed weekly cadence: one operations meeting where each decision-owner reports against their numbers — bids priced and hit rate, change orders captured and outstanding, cash position against the 13-week forecast, schedule risks. Thirty focused minutes with a scorecard beats forty hallway interruptions, and it trains your leaders to arrive with answers instead of questions.

The metric that tells you it's working: count the decisions that touch you per week, and watch it fall. The vacation test is the final exam — ten days genuinely unreachable, and the honest postmortem of what broke. Whatever failed is your next delegation project; whatever ran smoothly is now actually delegated. Owners who pass that test discover something disorienting: the company runs, and their job has quietly changed from making every call to building the people and systems that make the calls. That's not losing control. That's the first time you've actually had it.

Guard the rhythm from yourself, too. The failure mode after a good quarter is the owner drifting back into decisions that were handed off — answering the supplier call, repricing the change order — because being needed feels productive. Every reclaimed decision teaches the team the system was temporary. When someone brings you a decision that is theirs, the discipline is one question: “what does the standard say?”

The owner bottleneck is what got you here refusing to be what gets you there. The escape is mechanical, not motivational: define the standards you've been carrying in your head, delegate decision types with dollar thresholds instead of tasks, hold the line through the J-curve, and replace your open door with a weekly scorecard rhythm. Every piece of it compounds — each documented standard makes the next handoff easier, and each developed decision-maker develops the next one.

Start with the ten-day decision log this month. You'll find the not-actually-decisions pile within a week, and killing it costs nothing but honesty. Then hand off one threshold category — change orders under a limit is usually the right first move, because the capture system probably already exists on paper. The goal isn't for the company to need you less. It's for the company to be worth more than your calendar — which, someday, at exit or succession, is the only version anyone can buy.