Why Every Decision Still Finds Its Way Back to You

It’s 9:40 on a Sunday night. The house is finally down, the dishwasher is running, and you open the laptop for what you tell yourself will be ten minutes. Eleven threads are waiting. Every one of them is a question that stopped moving on Thursday because it was waiting for you.

Should we honor last year’s pricing on the renewal? Can the release slip a week? Do we backfill the support lead now or hold the seat until Q3? Is it fine to comp travel for the conference?

None of these are hard questions. That’s the part that should bother you.

You’ll answer all eleven before midnight, because you’re good at this and because the alternative is that nothing moves Monday morning. And next Sunday there will be eleven more.

I want to name what’s actually happening, because most founders misread it. This is not a workload problem. It isn’t even a delegation problem, at least not the way that word usually gets used. It’s a design problem. Somewhere along the way, your company’s decision system became a single node, and the node is you.

It made sense once. At eight people, routing every decision through the founder was the correct architecture. You had the most context, the most at stake, and the shortest path to the customer. A question asked at 10 a.m. was answered by 10:15. Centralization was fast.

At forty people, the same architecture is still running. Nobody changed it because nobody designed it in the first place. It was simply how things worked, and it kept working right up until it didn’t. The math changed underneath it. There are now more decisions per week than one person can hold, and each one waits in a queue behind all the others.

That queue has three costs, and only one of them shows up where you can see it.

The first is latency. A pricing question that waits four days for your attention isn’t a pricing question anymore. It’s a stalled deal, a rep who stopped pushing, a prospect who took a second call with someone else. Work doesn’t pause politely while it waits for a decision. It decays.

The second cost is what your waiting room does to your people. Watch what happens to a capable operator whose recommendations keep getting rerouted through the founder. The first few times, she argues her case. Then she starts pre-clearing things informally, catching you in Slack before committing to anything. Eventually she stops bringing recommendations and starts bringing questions. Not because she’s weak. Because she’s smart, and she has correctly learned what the system rewards. You didn’t hire passive people. You built a system that trains initiative out of them, one overridden call at a time.

The third cost is you. Your calendar is the most constrained resource in the company, and it is being spent on five-hundred-dollar decisions. The three or four calls this year that genuinely require you, the market bet, the senior hire, the product you finally kill, get whatever attention is left at 9:40 on a Sunday. That’s the inversion nobody books on a P&L: the company’s most expensive judgment applied to its cheapest problems.

Here is the reframe you're owed, because founders tend to convert this into guilt. You did nothing wrong. The routing that exhausts you today is the same routing that got you to eight figures. Growth didn’t break your judgment. It broke the assumption that your judgment could be everywhere at once.

So what actually fixes it? Not a delegation speech. I’ve watched leaders give the “I need you all to take more ownership” talk with complete sincerity, and I’ve watched the same decisions land back on their desk within three weeks. A speech doesn’t change anything because the system that routes decisions upward is still fully intact. People escalate because escalating is the rational move: it’s how things get decided here.

The fix is structural. Decide, explicitly, which decisions belong to which seats. Set thresholds: below this dollar amount, this risk level, this customer tier, the seat decides and informs you afterward. Write it down where everyone can see it. Then comes the hard part, which is holding the line when someone escalates a decision they own. The first time you send one back with “that’s your call, and I’ll back whatever you decide,” you’ll feel the whole company flinch. Do it anyway.

The first month is uncomfortable. Someone will make a call you’d have made differently, and once in a while they’ll be wrong. That is not the system failing. That is the tuition. The alternative is a company where nobody but you ever gets reps making real decisions, which means nobody but you ever gets good at it, which means the queue never shrinks.

I spent three tours as Chief of Staff to senior technology executives, inside organizations orders of magnitude larger than yours. The furniture was different. The failure was identical: work waiting on a single overloaded decision point while capable people stood by. The mechanics scale down to a forty-person software company just fine. The physics don’t change at all.

Your company shouldn’t need you in every decision. It needs you for the handful of decisions no one else can make, and it can’t have that version of you while you’re clearing eleven threads at midnight.

This is the first thing I look for when a founder asks me why everything still runs through them.