Agency owner stuck working in the business: how to find the seat you have not filled

You sell the account, you brief the team, you are copied on the escalation, and when the deliverable comes back short you fix it yourself at 9pm because that is faster than explaining it again. The agency makes money. It also stops the day you stop, and you have known that for about two years.

The advice you have already heard is to delegate more and work on the business instead of in it. That advice fails because it treats the problem as a habit, when the problem is structural: there is a seat in the company that nobody sits in except you, and until it has an owner, every task that belongs to it will find its way back to your desk.

Why the founder ends up holding everything

An agency under about $100k a month is usually built around the founder's skill. Clients bought you, the team learned from you, and the processes exist in your head because writing them down was slower than doing the work. That is fine at eight people and it fails at fifteen, because the number of decisions grows faster than the headcount does.

ZeroTo1 is the clearest case we have. The work was excellent, and the business only moved when the founders pushed it. Every hire, every client escalation and every process question routed back through them. They were also 100% referral dependent, which meant new referrals roughly matched churn and the company sat at $78k a month feeling plateaued. The real problem was structural rather than a lack of effort, and effort was the one thing the founders had already maxed out.

The three-week log that shows which seat is empty

Before changing anything, log every working hour for three weeks. Use a spreadsheet with four columns: selling, delivering, escalations, and running the company. Escalations means any hour spent on a client problem that someone else was nominally responsible for. Running the company means hiring, finance, and planning, which is the work the owner is supposed to be doing.

Most founders who do this honestly find that selling and escalations take 60% or more of the week, and running the company takes under 10%. The split tells you which seat is empty. If escalations dominate, the missing seat is a delivery lead or head of client services. If selling dominates, the missing seat is a salesperson, and the founder-led sales handoff covers how to make that transfer without losing close rate.

This is the same logic as the binding constraint: one thing limits the business at a time, and fixing anything else is wasted work that feels productive. Four candidates bind an agency: acquisition, conversion, delivery capacity, and margin. The founder's hours are the clue to which one it is right now.

Fill one seat, in a specific order

The instinct is to hire an operations manager to take everything off your plate at once. That hire fails more often than it works, because the role has no definition beyond "the things the founder does not want to do", and nobody can succeed in a seat shaped like that.

Fill the seat the log pointed at, and only that one. Give it three things on day one: a written list of the decisions it owns without asking you, a number it reports every week, and a weekly meeting where it reports that number out loud. The operating cadence article describes the meeting structure. Without the cadence, the seat exists on an org chart and the escalations keep arriving in your inbox anyway.

For ZeroTo1 the first build was management architecture: every role mapped, reporting lines drawn, a KPI per seat, and structured one-to-ones. Process codification came second, moving SOPs and the client management hub into ClickUp so a qualified hire contributed within weeks instead of shadowing the founders for months. The go-to-market system came after that, once delivery could absorb what sales would bring in.

The test that tells you it worked

Take two weeks off with your phone in a drawer and see what breaks. Most owners cannot bring themselves to run that test, so run the cheaper version: count the hours in the escalations column again, ninety days after the seat is filled. If the number has not halved, one of three things is wrong. The person in the seat lacks authority to make decisions, so the team still routes around them to you. The decisions were never written down, so nobody knows what the seat owns. Or you are still answering the messages, which trains everyone to keep sending them.

The third one is the usual culprit, and the fix is a rule rather than a hire: any client message that belongs to the filled seat gets forwarded to it with no reply from you, every time, for a month. The team learns within a fortnight that the shortest path to an answer no longer runs through the founder.

What it looks like when it works

ZeroTo1 went from $78k to $414k a month over 36 months, with headcount from 11 to 75 and EBITDA margin held while revenue grew five times. The founders still work in the business every week, on the parts of it that only they can do. Fraggell Productions is the same shape at a smaller scale: a pod-lead layer put between the founders and daily production took the company from £23k to £110k a month over 24 months, and the founders came out of the day-to-day.

If you want to know which seat is empty in your own agency, the diagnostic asks the questions the three-week log answers, and a call with us goes through the result.

Read next

• Getting the founder out of agency sales without the pipeline collapsing
• The agency operating cadence: what a weekly rhythm actually looks like