In a recent article on what an accountant actually does we described the most common frustration in small business: owners who like their accountant but wish there was more. More insight, more foresight, more of a feel for how the business is actually doing. We made the case that this “more” is a second job entirely, the job of a finance function rather than an accountancy service.
Here’s the part that might surprise you: that second job doesn’t begin with dashboards, forecasts or a fractional FD. It begins somewhere far less glamorous – in the bookkeeping.
Not whether it’s done, but how it’s designed. Because the difference between numbers that merely get filed and numbers that can run a business is decided at the categorisation level, transaction by transaction.
Two standards of bookkeeping, one word for both
Compliance needs surprisingly little from your books. For the annual accounts, the detail mostly rolls up anyway: sales is sales, wages are wages, and as long as everything is captured, supported and reconciled, the year-end will balance and the returns will be right. Bookkeeping done to this standard is not wrong. It’s doing exactly the job it was asked to do.
Management information needs something else altogether. It needs your numbers to answer questions like: which client is profitable, which service line is growing, what did that project really cost, is this month genuinely better than last month or does it just look that way? None of those answers exists at the “sales is sales” level. They exist only if the detail was captured on the way in.
Both standards are called bookkeeping, but they’re not the same product.
Where the management information actually comes from
The structure of your books should be a deliberate design, not a template. Each of your transactions gets coded to one of the ‘accounts’ in your books, and collectively this structure is called your ‘chart of accounts’. A chart designed for management splits revenue and costs the way you make decisions: by client, project, programme, site or service line, etc. An agency that codes by project can see project margin. One that doesn’t, can’t, and no amount of clever reporting later will recover detail that was never recorded. For the year-end filing, that whole structure summarises into a handful of statutory lines, disguising the detail that you use to make business decisions month by month.
Consistency is what makes comparison true. Management accounts live on comparatives: this month against last month, against budget, against forecast, against last year, etc. Those comparisons only mean something if the same things are coded the same way every time. Consistent coding is invisible in the annual accounts and absolutely everything in the monthly ones.
Journals are what make a month honest. Here is the least appreciated work in finance. A raw ledger lies about months. The annual insurance bill lands in January and ruins it; the big invoice raised early flatters March; the software renewal paid yearly makes one month look worse than eleven others. Journals fix this: prepayments spread costs across the periods they belong to, accruals recognise costs incurred but not yet billed, deferred income holds fees back to the terms or months they were earned in, payroll journals put the true staff cost in the right place, and monthly depreciation stops equipment distorting anything. Not one of these journals is needed until year-end for compliance. Every one of them is needed monthly for the truth. A monthly P&L without them isn’t management information.
And then there’s WIP, the exception that proves the rule. Work in progress (WIP) is the one place this discipline and compliance genuinely meet, because the year-end accounts do require it. But a business that only values WIP once a year, for the accountant and HMRC, is using a management tool as a filing chore. Valued monthly, WIP (and its construction cousins: applications, valuations and retentions) is what tells a project business its real position, which jobs are quietly going wrong, and what this month actually earned. Same calculation, entirely different purpose.
Why this explains the frustration
Now the recent article‘s puzzle resolves neatly. When owners ask their accountant for more insight, the honest obstacle is often not willingness or even skillset. It’s that the ledger was never built to answer the questions. Books maintained to compliance standard simply don’t contain client margins, true monthly costs or project positions. The accountant would have to rebuild the data before they could analyse it, and nobody’s fee covers that.
Which is also why bolting an FD or a reporting tool on top of compliance-grade books disappoints. The steering layer can only be as good as the sensor layer. Garbage in, dashboard out.
One team, designing the books for their real job
This is the quiet logic of bookkeeping and finance done together, and it’s exactly how GrowBe works.
Our bookkeeping isn’t a separate cheap layer done to filing standard and handed upwards. It’s designed from day one by the same team that will produce your management accounts, run your forecasts and sit with you every month: the chart of accounts built around your decisions, the journals posted monthly so every month is true, WIP and deferred income treated as management tools that happen to satisfy the year-end too. The compliance then falls out of the side, almost for free, because books that are true every month are trivially true at year-end.
If you want the “more”, this is where it starts. Not with a smarter report on top of the same old ledger, but with a ledger built by people who intend to use it.
