An agency, consultancy or professional services firm has one dominant cost — people — and one dominant question: which work is actually worth doing. Standard bookkeeping answers neither, because it wasn't built to.
The default chart of accounts is organised around the tax return. Revenue lands in one or two lines. Salaries land in another. Everything is accurate, nothing is decomposable, and the numbers that would tell you something arrive averaged into oblivion.
If retainer work, project work and one-off advisory all land in "Revenue," you cannot see that one of them carries half the margin of the others. Most firms have a service line quietly subsidised by another, and most owners can't name which without a manual exercise.
When every salary sits in one line, you can't separate the cost of delivering the work from the cost of running the firm. Gross margin becomes unknowable — and gross margin is the number that tells you whether the model works.
Client profitability requires revenue and delivery cost tracked at the client or project level. Without it, the largest client looks like the best client, which is true remarkably often and false often enough to matter.
Utilisation is the central operating metric of a services business, and it appears nowhere in standard accounting. Growth constrained by demand and growth constrained by capacity are opposite problems, and you can't tell them apart from a P&L.
The fix is structural and you only do it once. Four changes carry most of the value:
Books organised for the tax return tell you what you owe. Books organised around your operations tell you what to do.
Gross margin by service line. Revenue minus delivery cost, per line. If one is below the others by more than a few points, either the pricing is wrong or the delivery is inefficient — and those have different fixes.
Utilisation. Billable hours over available hours. Most firms target 65–75% for delivery staff. Below the range, you have capacity you're paying for. Above it, you're heading for quality problems and attrition.
Realisation. What you actually collected against what you could have billed at standard rates. The gap is discounting, scope creep and write-offs — and it's usually larger than owners expect, because each individual instance felt reasonable at the time.
The one technical point worth getting right. If you deliver in March and bill in April, cash-basis books show a bad March and a great April, and neither reflects what happened.
For firms of any size, accrual treatment with proper WIP tracking is the only way the monthly numbers mean anything. It also matters enormously if you ever sell — buyers discount earnings they can't tie to a period.
Restructuring a chart of accounts and rebuilding a close process is typically a one-off project measured in weeks, not an ongoing expense. The output is that every month afterwards produces answers rather than records.
If you're currently running an agency or consultancy on books that show one revenue line and one payroll line, that project is almost certainly the highest-return finance work available to you.