How agencies actually make money

The most reliable route from a wage to a business, and the one most honest about being work. The economics come down to utilisation and the gap between what you charge and what you pay.

One of the common ways of making money, looked at closely.

Of the six, this is the least glamorous and the most reliable. It is also the one whose economics are simplest to state, which makes it a good lens for the others.

The arithmetic

You charge a client £100 an hour for work. You pay someone £40 an hour to do it. The £60 covers sales, management, tools, premises, the hours nobody billed — and what remains is profit.

Two numbers decide everything.

The rate gap — what you charge minus what you pay. This is set by how hard the client’s problem is to solve elsewhere, not by how long the work takes.

Utilisation — the share of paid hours that are actually billed. This is the number that quietly kills agencies. Staff are paid for every hour; only billed hours produce revenue. Time spent on sales, admin, rework and gaps between projects is paid and unbilled.

An agency at 80% utilisation and one at 55% can have identical rates and identical staff, and only one of them is a business.

Which channel this really is

It is the wage channel, enlarged.

You have escaped the cap on your own hours by employing others — genuinely important, and the reason an agency scales where freelancing does not. But the business still runs on labour. Revenue is a function of people employed and hours billed. Stop selling and it stops.

That is not a criticism. It is why an agency is achievable: it needs no capital, no audience, and no invention. It needs a client with a problem and the ability to solve it.

What is genuinely owned

Three things, and they are what make an agency saleable where a freelance practice usually is not.

Client relationships, particularly retained ones. Recurring revenue from clients who stay is the difference between a business and a series of projects.

Reputation in a niche. “The firm that does X for Y-type companies” is an asset. “A digital agency” is not — it competes with everyone on price.

A method that belongs to the firm. If the way work gets done is documented, repeatable and independent of any individual, the firm has something. If the quality lives in one person’s head, the firm is that person.

What the evidence says

Better than most on this list, because agencies are registered businesses that file accounts.

Survival resembles small businesses generally — most do not reach a decade. The failure modes are well known and boringly consistent:

Client concentration. One client at 40% of revenue is not a client, it is an employer who can fire you without notice or severance.

Cash flow. Staff are paid monthly; clients pay on 30, 60 or 90-day terms, sometimes late. Profitable agencies fail from timing.

Founder dependence. If the founder sells every deal and reviews every deliverable, growth stops at their capacity and the firm cannot be sold.

Note that none of these is lack of demand. Agencies rarely die because nobody wanted the work.

How the top firms differ

The change is from selling hours to selling outcomes.

Narrow rather than broad. A specific problem for a specific industry supports higher rates, shorter sales cycles and referrals, because expertise is legible. Generalists compete on price against everyone.

Price the outcome, not the time. If the work saves a client £500,000, the fee relates to that, not to the days it took. This decouples revenue from hours — the only way out of the utilisation trap.

Productise the repeatable part. Once a method exists, some of it becomes a fixed-scope offering with predictable delivery. Margins improve because the work is known.

Build recurring revenue. Retainers and managed services change the business from hunting to farming, and are what buyers pay a premium for.

What AI changes

The pressure is real and specific, and it is not on the whole business.

It compresses the cost of producing the deliverable — the copy, the design, the code, the analysis. If that is what you sell, your price falls with your cost, and clients will eventually notice.

What it does not compress is deciding what should be done, and being accountable for it. A client who does not know which thing to build gains nothing from cheaper building. Someone who understands their situation, recommends a course of action, and can be held responsible when it is wrong is providing something a tool does not.

So the shift is from execution toward judgement, which is the same direction the top firms were already moving for their own reasons.

If you are considering it

Sell before you hire. Demand first, capacity second. The reverse is how utilisation collapses.

Watch client concentration from day one. No client above roughly a quarter of revenue, and treat approaching that as an emergency rather than a success.

Model cash flow, not just profit. Know how long you can pay staff while a large invoice is late, because it will be.

Write the method down as you go. It is the difference between building a firm and building yourself a demanding job.

And be clear it is work. This is the option on the list that most reliably produces income and least resembles passive anything. For many people that is exactly the right trade — it funds everything in stage 3 — but it should be chosen knowingly.