Bookkeeper for Marketing Agencies

How does money flow into an agency?

An agency is paid one of three ways: a flat monthly retainer, a one-time project fee, or a base fee plus a commission on results.

The retainer is the most common. You bill the client on the first of the month and the money lands right away, but you have not earned it yet, because the work spans the whole month. If the book records the whole payment as income on the day it arrives, the first week of the month looks great and the rest of the month looks empty. A clean set of books keeps that payment in a deferred revenue line and releases it as the work gets done, so profit shows up across the month instead of in one burst.

A project is the same issue at a bigger scale. A typical project is billed 50% up front and 50% on delivery. That opening payment can sit in deferred revenue for months. If you record it as income the month the contract is signed, the P&L shows a big jump that month, an empty stretch while you do the work, and a drop when the final invoice lands. That is not a picture you can manage from.

A performance fee is the opposite problem. When part of your pay depends on results, what you get is decided by the client's attribution dashboard, not by anything you control. You cannot predict it and you cannot book a guess. The clean way is to record the commission only once it is confirmed.

So in all three patterns, cash in the bank and revenue on the books are running on different timetables. If the books treat them as the same thing, your reports will swing up and down with payment dates instead of showing the real shape of the work.

What does an agency actually spend money on?

The money goes to team time, media, software, and the cost of bringing a client on board, and two of those are money that passes through your hands rather than being yours.

Team time is the biggest cost at most agencies. You are paying for the hours that research, build, manage, and report. Agencies commonly work on a markup of 10% to 20% over what that work costs, which is what makes labor costs the center of the whole margin story.

Media is the place where the books get confused with reality. When you buy a client's ad spend, the whole budget moves through your bank account on its way to the ad platform. The fee you add is typically 3% to 15% of that spend, and that fee is your revenue. The 80% of a $20,000 media bill passes through. If the whole $20,000 is recorded as revenue, your top line inflates, and, because you also recorded the $20,000 going out to the platform, the margin drowns.

Software splits the same way. The tools you use to run the agency, like reporting dashboards and design software, are a normal business cost. The client's own software that you license and manage on their behalf is another pass-through, like ad spend. The invoice for that license may land in your account and go right back out, and none of that is revenue you get to keep.

And every new client has a cost before the first retainer. Discovery and onboarding work is a real set of hours. Pricing guides put that work in the $10,000 to $20,000 range for a local client, more for a national one, and it is often not billed separately. It gets worked into the ongoing fee. A clean set of books tracks that time as part of the client's true cost instead of letting it hide in a general labor line.

What goes wrong in the books?

The failures repeat: the books are only touched at tax time, revenue is recorded when the payment lands instead of when the work is done, and pass-through money sits in revenue as if it were yours.

• The bookkeeping role stays empty while the shop grows. Agencies are started by creative and technical people, and neither partner wants the ledgers. So the books get opened once a year for the tax deadline, and the only number the owner ever gets is the tax due. Nobody sees the month-by-month profit until it is too late to do anything about it.
• Revenue gets recorded when the invoice is paid, not when the work is earned. That means a $4,000 retainer and a $20,000 ad budget create a great-looking month when the deposits clear, and a terrible looking month later when the work shows on the other side. The owner fights a whipsaw that is only a timing problem, but it gets treated as a performance problem.
• Pass-through money sits in revenue. Ad spend and licensed software inflate the top line and the tax position, and the real margin hides underneath. This is the single most common mistake I find in agency books.
• There is no cost per client. The P&L lists broad expenses, but nothing says what it costs to serve any one client. So there is no way to know if a retainer is worth carrying, which client to raise, and where the profit actually lives.

The outcome shows up as a cash fight. Receivables grow, payables grow, and the owner is forever wondering why there is no money even though the invoices look fine. The books are not giving them the reason, because the actual problem is earlier: revenue is being recognized at the wrong time and pass-throughs are being counted as income.

What should have its own line in the books?

Each service line and each client needs its own numbers in the books, and pass-through money needs a separate track that does not look like revenue.

Service lines matter because they are different businesses. A monthly retainer for ongoing work, a one-time website rebuild, and a campaign with media behind it each have their own margin and behave differently as you grow. If the books lump them into one revenue number, you cannot decide which one to push and which to cut. To make that call you need cost and revenue per service, not one blended picture.

Per client, the number you need is cost to serve. Owners in agency forums ask specifically for cost per client acquisition and cost to service a client in a given tier of service. That cost is the team hours on that client, the share of software, and the share of ads or onboarding you carry. With that number, a $2,000 retainer that costs $1,400 to serve is a real decision, and a $2,000 retainer that costs $2,100 is a loss you can see and raise or cancel.

Pass-through money gets its own treatment. The clean way is a clearing account that the client funds, and the agency pays out of when the media or the software bill arrives. When it flows through, it has invented display. At the end of the month the account balances to zero, and none of that pass-through has ever touched your revenue or your margin.

Which numbers matter, and what do they change?

Four numbers do the job: earned revenue, margin per client and service, cash you actually hold in reserve, and how much of the team's capacity is already sold.

Earned revenue is the accrual number. It counts revenue in the month the work is earned, not the month the invoice is paid. That is the number that tells you whether the business actually made money last month, because the cash and the work rarely line up in this trade. An why you chose between two runs of the P&L, accrual and cash, and know which one answers which question.

Margin is the second. Different sources quote agency net margins in different ranges, anywhere from the low teens to the mid thirties, and the truth is that the sources do not agree. What that means is that no benchmark is more useful than your own book. When the books are clean and pass-throughs are out, your margin number is a guide for what to price and what to decline, and you do not need a trade average to tell you that.

A cash reserve is the third. Some advisors suggest holding about 10% to 30% of annual revenue in reserve to ride the months when clients are late or when a project wraps and the next one is still in the pipeline. That is a range of traffic. It tells you how a slow month affects you and whether you can afford to turn work down.

Capacity and utilization is the fourth. It is the percentage of the team's available posts who are actually billing. Low utilization means you have room to take on work or should be selling. High utilization means you are about to miss deadlines, and the decision is whether to hire or to raise prices. Clean labor, posted per client project, is what makes this number possible.

What decisions should the books support?

The books are only as good as the decisions they settle: pricing a retainer, deciding which client and which service line to keep, knowing when to hire, and choosing when to take the next project.

Pricing, the main one. If the cost to serve a client in a tier is $3,000 a month, then a $2,000 retainer is a loss you are carrying in place of the time, and the books should say it plainly. The cleanest agencies stop selling time and sell a fixed outcome or scope for a fixed fee. That only works when you know the cost under the fee. That is the decision the cost number serves.

The service mix decision follows from the margin per line. A service that is losing money because media or setup is under-priced is a candidate for a price change or a sell or not. A service that carries the whole company is one you protect and push. You cannot tell the two apart without line-level margin.

Hiring and selling decisions rest on utilization. At 70% sold, you take the next project and enjoy it. At 95% sold, taking the project means you pay a late or every job slips. Utilization is the number that carries the hire, not the and a single. The books show you the utilization so the decision stops being a guess.

And the project you turn down is a real decision too. A clean view of cost, margin and capacity turns from a case into a business. When a $10,000 project does not cover a pattern in the following client start, you can say no with the books behind you, and then you can argue about timing and price instead of guessing at profit.