Nearly every agency owner between five and twenty clients is the bottleneck. Work waits on them for approval, for the difficult client, for the thing only they know how to do.
They know this. They have tried to delegate. It went badly, someone had to fix it, and the conclusion was that it is faster to do it themselves — which is true in any single instance and false across a year.
Here is why it fails and what order to fix it in.
Why delegation fails
It is delegated as tasks rather than outcomes. "Write four posts for the dentist" makes you the planner forever. "Own the dentist's content calendar" transfers the thinking.
The standard is in your head. You know a post is wrong when you see it but have never written down what makes it right. Nobody can hit an unstated standard.
No decision authority. If every judgment call routes back to you, you have added a step rather than removed one.
Handed off during a crisis. Delegation is a training investment, and training during a fire fails. It has to start in a slow week.
Judged against your best work. A new coordinator will produce work worse than yours. The relevant question is whether it clears the bar the client is paying for, not whether it matches yours.
The order to delegate in
Sequence matters. Most owners start with the wrong things.
First: scheduling and publishing. Mechanical, verifiable, low judgment. Someone else moves approved content into the calendar. Immediate hours back, near-zero risk.
Second: reporting assembly. You still write the analysis. Someone else pulls numbers and builds the document. Reporting is one to three days a month at fifteen clients, and most of it is assembly.
Third: approval chasing. Following up on pending reviews requires persistence, not expertise. Give the sequence and the templates.
Fourth: content drafting. The real handoff. Start with one straightforward client, not your hardest one. You review everything for the first month, spot-check the second, review by exception after that.
Fifth: client communication. Routine email and updates first. Strategy calls last. Do not start here even though it consumes the most time, because trust in the account has to be established first.
Last, or never: strategy and difficult clients. Some owners keep these permanently. That is a legitimate choice as long as it is a choice.
Write the standard down
Before delegating drafting, write a one-page document per client containing:
- Three voice adjectives with two example sentences each
- Two sentences that would be wrong, and why
- Off-limits topics
- Who approves and what happens when they are away
- The four content pillars
- Three examples of posts that performed well and why
One page. Longer than that and nobody reads it.
The test: hand it to someone who has never worked on the account and see if they can produce something usable. If they cannot, the document is missing something — and the missing thing is what you have been holding in your head.
Grant decision authority explicitly
Write down what they decide without asking:
Decide alone: wording, image selection within the brand set, posting time within the agreed window, routine comment replies, minor revisions from client feedback.
Ask first: anything touching pricing or offers, anything about a competitor, response to a negative comment, scope changes, anything they find risky.
Ambiguity defaults to asking, which defaults to you being the bottleneck again.
The quality bar
Set the bar as "the client would approve this," not "I would have written this."
Track approval rate. If a coordinator's work is approved at 90 percent or better, the delegation is working regardless of whether you would have phrased things differently.
If it is below 70, the problem is usually the voice document rather than the person. Look there first.
The first month costs you time
Be clear-eyed: delegating drafting costs you more hours in month one than doing it yourself. Reviewing, correcting, and explaining is slower than writing.
Month two is roughly break-even. Month three is where you get the hours back, and they keep coming.
Owners who abandon delegation almost always abandon it in month one, at the point of maximum cost and zero return.
What to do with the reclaimed time
The failure mode is filling it with more client work, which recreates the bottleneck at a higher client count.
Spend it on: selling, raising prices, building the systems that make the next delegation easier, and taking actual time off so the business is not running on your endurance.
The test of whether it worked
Take a week off with your phone off. Not a working holiday — genuinely off.
If the agency runs, it is a business. If it does not, it is a job that employs other people, and you have identified exactly what to fix next.