A junior runs your agency account because the economics leave no other option. Agencies target 50-60% gross margin. Half your retainer is spoken for before anyone works an hour, and the hours that remain have to come from the cheapest people on payroll.

That is a structural fact rather than a scandal, and you can contract around it if you know what to ask.

There is a genre of founder post that recurs on LinkedIn every few weeks. The shape barely varies: we fired our agency this morning, and here is what I found when I finally looked at who was working the account.

The owner who pitched us never touched it again. A coordinator two years out of school ran strategy, execution, and reporting. When performance dropped, nobody senior even noticed.

The founders who write these posts usually blame themselves for picking badly. Then they pick again, more carefully, and eighteen months later the new agency has drifted to the same staffing.

I have sat on the buying side of this table as a CMO, hiring and firing agencies with real budgets behind the decision. The pattern stopped surprising me once I priced out what a retainer can mathematically buy.

Why is a junior running your account?

Start from numbers agencies publish about themselves. Industry benchmarks put a healthy agency at a gross margin of 50-60%, with overhead running near 30% of gross income and 15-20% surviving as net profit.

None of that is greed. An agency below those numbers dies in the next soft quarter, and everyone who has watched one die runs the survivors harder.

But hold the retainer you pay up against those targets and the consequence falls out on its own.

An illustrative $10,000 retainer, decomposed Senior strategy and review: about $1,100 The hours the pitch deck showed you Delivery hours: about $4,300 total The only money that touches your account, mostly billed by the most junior hands Overhead: about $2,900 Office, software, management, and the non-billable half of the org Profit: about $1,700 The margin the model exists to protect Illustrative split, consistent with published agency benchmarks: 50-60% gross margin, overhead near 30% of gross income, 15-20% net.
The decomposition is the answer to a question founders rarely ask at the pitch: of this money, how much converts into hours worked on my account, and whose hours are they?

Where does a $10,000 retainer actually go?

Run an illustrative $10,000 a month through the benchmark targets. Roughly $4,300 can become delivery: hours worked on your account. About $2,900 covers the building, the software, the managers, and the non-billable half of the organization. About $1,700 is the profit the whole model exists to protect.

Now ask whose hours the $4,300 buys.

A senior strategist carries a loaded cost somewhere near $120 an hour. Staff your account entirely with people like that and the budget buys about 35 hours a month. One working week, for a retainer that runs $120,000 a year.

A retainer that delivers one quiet week of work per month feels like neglect, and clients leave over it. So the mix shifts to where the hours are: a coordinator at a $35 loaded cost delivers 120 visible hours for the same budget.

What $4,300 of delivery buys, in hours All senior about 35 hours: one working week Typical mix 10 senior + 80 junior All junior about 120 hours Loaded senior cost near $120 an hour; junior near $35. Both illustrative. A retainer must produce enough visible hours to feel like service. Only the junior-heavy rows can do that inside the margin target, which is why the mix drifts down the moment the contract is signed.
No villain required. This is the only allocation that clears the benchmark margins and still fills the month with enough visible work to feel like service.
Nobody decided to shortchange your account. The staffing mix is the only one that satisfies the margin target, the hour count, and payroll at once.

The pitch team and the delivery team

The same arithmetic explains the bait-and-switch feeling of the sales process, without requiring anyone to have lied to you.

A senior person's hour has two possible uses. Spent on your account, it produces $120 of delivered value against a budget that cannot afford many of them. Spent in a pitch, it can close a $120,000 contract.

Every rational agency routes senior time toward the second use. Seniors sell, juniors deliver, and the org chart is built around the routing: benchmark utilization targets expect delivery staff to run 75-85% billable, while the principals who impressed you in the pitch barely bill at all.

So the person who diagnosed your funnel in the sales call was never going to be the person watching it. The diagnosis was the product being sold. The watching is the product being delivered, by whoever the margin allows.

How can you tell who will actually work the account?

The defense is contract language rather than better vibes in the pitch. Three questions, asked in writing, before signing.

Three questions to ask before signing ASK A GOOD ANSWER THE TELL 1 Who is my delivery team, by name, at how many hours each per month? Names and numbers, in the contract "A senior point of contact" 2 Who attends the weekly after month one? The same names, and the contract says so A rotating account manager 3 What share of hours on my tier is worked by people with 5+ years in the channel? A percentage they can defend A tour of the leadership page Every answer exists in the agency's own resourcing system. An agency that will not put it in the contract has answered anyway. Ask in writing, before the pitch team leaves the room.
None of the three questions is hostile. Agencies that staff honestly answer them in a sentence and tend to raise the subject first.

Agencies that staff honestly answer these in a sentence and often raise the subject themselves, because their staffing is a selling point against everyone else's.

An agency that offers a senior point of contact in place of named hours has also answered. A point of contact is a routing address. You are asking who does the thinking.

One more test costs nothing: in month two, note who speaks in the weekly call. The org chart you are actually paying for introduces itself within six weeks.

When an agency is still the right answer

None of this math argues against agencies. It argues for buying what the model is built to sell.

The model is excellent at execution capacity: producing ad creative at volume, running a large paid program's daily operations, covering a specialist channel you cannot justify hiring for. Hands on day one, scaled up and down at will. No individual hire competes with that.

What the model structurally cannot sell you is a senior person who holds your revenue number in their head between calls. The margin math spreads senior judgment across dozens of accounts, and judgment spread that thin arrives quarterly, in a deck.

Buy execution from agencies. Source judgment separately, and make whoever provides it accountable for the number itself.

When I audit a company's vendor stack, the retainer decomposition above is usually the first exercise, because it predicts everything else we find. Let's talk.