Why agencies are bad at marketing themselves

Agencies sell growth for a living and then grow by referral. Every owner in the industry knows this and most treat it as a joke about cobblers and shoes. It's not a joke. It is a structural problem with a specific cause, and the cause is worth understanding because it also explains the fix.

Your own account is the only one with no client

An agency runs on capacity. Every hour of senior attention is allocated somewhere, and every allocation has a client attached who will notice if it slips. Except one. The agency's own marketing has no client. Nobody emails on Tuesday asking where the campaign is. There's no invoice at risk, no status call, no relationship that gets damaged. So when the week compresses, and it always compresses, the account that gets sacrificed is the one that can't complain.

This isn't weak discipline. It's a rational allocation of scarce attention under pressure, repeated every week for years, and it produces a business that's excellent at growing other companies and structurally incapable of growing itself.

The size of that competition is measurable. When Kontext moved its own growth work out of the internal pool, Tamas got back twelve to sixteen hours a week. That is not a productivity anecdote, it's the number of hours the agency's own account had been losing to client delivery every single week, which is why it never got done.

The three symptoms

You can tell how far along this is from three things.

Your pipeline is entirely warm. Every deal traces back to somebody who already knew somebody. That feels healthy right up to the quarter when the network is quiet, and then there is nothing behind it, because nothing was ever built behind it.

Your own site is two years old. You would not let a client ship what is currently on your homepage. The case studies stop at a project you finished in 2024. Everyone in the building knows and nobody has the hours.

Proposals are the bottleneck, not leads. Interest exists. What does not exist is the two hours to turn interest into a written proposal, so opportunities queue behind delivery and go cold in the queue.

That third symptom is the one owners underestimate. It looks like a demand problem and it is a throughput problem, which means spending more on demand makes it worse.

Why the usual fixes do not hold

Agencies try three things, in roughly this order.

Block time. Friday afternoons are for our own marketing. This works for about five weeks. Then a client escalation lands on a Friday, and the block is the obvious thing to move, and it never comes back.

Give it to a junior. The person with the most available hours has the least context about the offer, the market, and why clients actually buy. Output appears, quality is thin, and the owner ends up rewriting it, which costs more time than doing it.

Hire a growth person. This can work. It is also a real salary against a function that will take months to produce, in a business whose cash flow is lumpy, and it puts the entire capability in one person who can resign.

All three fail the same test: they still depend on someone inside the building having hours. Anything that depends on internal capacity will be sacrificed to client work, because client work is what pays.

What actually changes it

The structural fix is to move your own growth work out of the pool that client work draws from. Not to find discipline. To remove the competition for hours entirely.

Tamas at Kontext Group is the example we can talk about in public. His agency went from one or two proposals a week to more than fifteen, recovered twelve to sixteen hours a week of his own time, and profit was up 31 percent. The interesting part is the order of those three. The hours came back first, and the proposals followed, because the proposals were never blocked on demand.

What made that possible was not another dashboard. Revenue Engine's AI kept the recurring work moving without anyone at the agency having to remember it: the list stayed current, outreach went out on a cadence, replies were worked and recorded, and proposals stopped queuing. Kontext chose the goals, voice, boundaries, and review points. The AI kept the cadence.

The test worth running

Go back through your last twelve months of new business and label each win by source. Referral, network, inbound from something you published, outbound you initiated, or repeat.

Then look at the outbound column. If it is empty or nearly empty, you do not have a demand problem to solve later. You have a single point of failure, and it is your personal network, and it is the one asset in the business you cannot scale by working harder.

You can also put numbers on the throughput side with the agency cost calculator, which prices what the current setup costs against what it produces.

Where to go next

If this is your business, the agencies page lays out what Revenue Engine runs for agencies specifically and what changes in the first months. It is also worth reading the eight programs, because the diagnosis for most agencies comes back the same way: strong offer, strong delivery, no Reach, no Control.

Run the twelve-month source test tonight; it takes twenty minutes and you already have the data. If the outbound column comes back empty, the agency intake is the fastest way to get a read on your firm specifically rather than on agencies in general.

Get my growth plan