Management accounts for agencies: what a proper monthly pack tells you.

There are two kinds of accounts your agency will ever produce. One is a legal obligation, filed once a year, read by almost nobody. The other has no legal existence at all, and it is the single most useful document in the business. Most agency owners have only ever seen the first kind, which is why the word accounts makes them think of compliance rather than decisions.

Management accounts are the numbers a business prepares for its own use, usually monthly, in whatever format actually helps the people running it. There is no template from HMRC, no filing deadline, no fine for skipping them. That freedom is the point, and also the problem. Because nobody checks, the quality ranges from a genuinely useful decision tool to a PDF of a profit and loss report that arrives late, says nothing, and gets deleted.

The gap is enormous and most owners never see across it. The SME Finance Monitor, a survey of tens of thousands of UK businesses, has consistently found that only around half of SMEs produce regular management accounts or keep a business plan, and the researchers noted the figure barely moves from year to year. Half of British businesses are steering on bank balance and instinct. This article is about what the other half get, and how to tell whether the pack you receive belongs in the first group or the second.

Two sets of accounts, two completely different jobs

Statutory accounts, sometimes called annual accounts, are the set the law requires. They exist for people outside the business: Companies House, HMRC, lenders, anyone checking the public record. The format is prescribed, the accounting rules are fixed, and the timetable is generous to the point of uselessness. A private company gets nine months from its year end to file, and the corporation tax bill lands nine months and one day after the period closes. Do the arithmetic on that. By the time a set of statutory accounts becomes public, the oldest transactions inside it can be twenty one months old.

Nobody runs a business on information that stale, in the same way nobody drives by studying where the car was last spring. Statutory accounts are the rear view mirror, and a legally mandatory one. They confirm where you have been. They are honest, standardised and completely silent on the only question that matters day to day, which is what you should do differently this month.

Management accounts are the steering wheel. They are built for the people inside the business, they arrive while the month they describe is still warm, and every section exists to set up a decision. Same underlying ledger, opposite purpose. If your accountant produces your statutory accounts and nothing else, you do not have a finance function. You have a filing service.

“Statutory accounts confirm where the business has been. Management accounts exist to change where it goes next.”

What a proper monthly pack contains

A real management pack is not one report but a short stack of them, each answering a different question. The test for every page is brutal and simple: what decision does this section set up? If a page cannot pass that test, it is decoration. Here is the anatomy of a pack that earns its place in your calendar.

The P&L, read against a plan

The profit and loss statement, the P&L, lists what the business earned, what it spent, and what was left. On its own it is a fact sheet. It becomes a management tool the moment you put two more columns next to it: what you budgeted for the month, and the difference between plan and reality. That difference is called a variance, and variances are where the useful information hides. A month that lands exactly on plan needs thirty seconds of your time. A month where contractor costs came in 40% over budget needs a conversation, this week, about whether a project is overrunning or a client has quietly expanded the scope of a retainer.

For the comparison to mean anything, the numbers have to be prepared on an accruals basis, which is accountant language for matching income and costs to the month the work happened rather than the month the money moved. A February invoice paid in April belongs to February. Without that matching, every month is distorted by payment timing and you end up investigating noise instead of signal. And one month is never enough on its own. A proper pack shows the trend, usually the last six or twelve months side by side, because a single bad month is often nothing while three gently worsening months in a row are always something.

Here is the sort of decision this section produces. Suppose salaries are on budget but freelancer spend has exceeded plan for three consecutive months, and the pipeline shows the work keeps coming. The numbers have just told you a permanent hire is cheaper than the rolling freelance bill, and told you three months earlier than your gut would have. That is the entire purpose of a variance column: it converts a vague feeling that things are busy into a costed decision with a date on it.

Margin by client, where the real story lives

Gross margin is what remains from your revenue after the direct costs of delivering the work, meaning the delivery team, the freelancers and the tools that only exist because clients do. It is the truest measure of whether the core of the agency works. And here is the uncomfortable part: the overall number can look fine while individual clients are underwater, because your best accounts are quietly subsidising your worst. Only a client by client breakdown shows you which is which.

The margin view also catches the slowest and most expensive failure in agency finance, the kind that never shows up in a revenue chart. Watch what happens to a business whose top line never moves.

Revenue in that picture is identical for six straight months, so a founder watching the top line sees a stable, healthy business. The margin line tells the truth: ten percentage points have evaporated since January, which on £80,000 a month is £8,000 of profit gone, every month, or close to £100,000 a year if nothing changes. The causes are always mundane. Two retainers picked up extra deliverables nobody repriced. A senior hire replaced a mid level leaver on the same accounts. A client that needed ten hours a week in January needs eighteen by June for the same fee. No single week felt like a crisis, which is exactly why only a monthly margin view catches it. The decision it drives is specific: identify the two accounts responsible, put a number on the scope creep, and open the repricing conversation this quarter instead of discovering the damage in next year's statutory accounts.

Cash, and the next thirteen weeks

Profit and cash are different things, and the difference is timing. The P&L records income when work is done. The bank account records it when clients actually pay, and UK clients are not in a hurry. Xero's Small Business Insights, drawn from anonymised records across more than 400,000 UK small businesses, put the average time to payment at 29 days in early 2026, with invoices settled around eight days after they were due. Meanwhile your costs are punctual to the day: payroll, rent and HMRC do not run eight days late. That mismatch is how a profitable agency ends up sweating a payroll run, and it is why the cash section of a pack matters more than any other on a bad month.

The centrepiece is a 13-week cash flow forecast, a week by week projection of money in and money out for the next quarter. Thirteen weeks is the standard horizon for a reason: long enough to see a VAT bill or a quiet patch coming, short enough that the estimates stay honest. A good one lists expected receipts by client and week, based on when they actually pay rather than when they are supposed to, and every committed outgoing including the tax already accruing. Alongside it sits an aged debtors report, which simply groups unpaid invoices by how overdue they are, so the chasing effort goes where the risk is.

The decision this section drives is about time. Suppose the forecast shows the balance dipping below your comfort level in week nine, because the quarterly VAT payment lands in the same week two large invoices are, on current behaviour, likely to still be unpaid. Seen nine weeks out, that is a to do list: bring the chasing forward, agree a payment date with the slow client, shift a discretionary purchase by a month. Seen the week it happens, the same event is a crisis with your name on it. Nothing about the cash changed. What changed is when you found out.

Pipeline, WIP and whether to hire

Everything so far describes money that has already been earned or spent. The forward half of the pack looks at what is coming. Pipeline is the revenue you have sold but not yet started plus the deals likely to close, weighted by how likely they really are. Work in progress, WIP, is the opposite corner: work you have already delivered but not yet invoiced. WIP deserves more fear than it gets, because every pound of it is effectively an interest free loan from you to your clients. An agency with £40,000 of unbilled work has £40,000 of its own cash parked in other people's businesses, usually because invoicing waits for month end or for someone to write up the time.

Set pipeline against your delivery capacity, meaning the hours or headcount you are already paying for, and the hardest decision in agency life becomes an arithmetic problem. If sold work covers 110% of capacity for the next quarter and the pipeline behind it is solid, the question is not whether to hire but how fast you can. If sold work covers 60% of capacity, a hire is not brave, it is a payroll cost in search of a job, and the honest move is to fix sales first. Owners without this page make the call on nerve and optimism. Owners with it make the call once, with numbers, and sleep afterwards.

A handful of KPIs, not a dashboard of forty

The last numerical page is a short list of key performance indicators, KPIs, which are simply the measures that drive everything above. For most agencies four are plenty: gross margin percentage, net profit margin, average debtor days, and revenue per head. Each one is an early warning for a different disease. Debtor days creeping from 30 to 45 predicts a cash squeeze months before it arrives. Revenue per head sliding while headcount grows says the team is expanding faster than the work. A wall of forty metrics is a substitute for thought. Four, tracked every month and compared with last month, is management.

Then comes the page that separates a real pack from a data dump: the commentary. Three or four sentences from whoever prepared the numbers, in plain English, saying what moved, why, and what they recommend doing about it. Then the actions, each with an owner and a date. If your pack has never once caused anyone to do anything, it is not a management pack. It is a subscription to a document.

Red flags that your pack is a formality

How to read the pack in twenty minutes

A pack built the way described here does not need an afternoon. It needs twenty minutes, in a fixed order, at the same point every month. The order matters: you read cash first because cash is the thing that kills quickly, and you read the backward looking pages after the forward looking ones have told you whether anything is urgent.

Do this with the person who prepared the numbers rather than alone, because the second best output of a management pack is the questions it makes you ask. Why did margin dip on that account? What happens to week nine if this invoice slips? Could we bring billing forward a week? None of those questions require financial training. They require current numbers and twenty minutes of honest attention, which is roughly what most people spend choosing something to watch in the evening.

The cadence is the part people underestimate. A pricing problem caught in month two costs you a month of margin. The same problem found at year end has run for twelve. Reviewed monthly, small course corrections compound in your favour; skipped for a quarter, small problems compound instead. Twenty minutes a month is the cheapest insurance a business can buy, and it is the difference between deciding with the numbers and being surprised by them. If you want a first data point while you get the pack itself sorted, the agency benchmarks scorecard compares six of your P&L numbers with other agencies in a couple of minutes.