I'm Damon Millar. I'm a UK accountant, fractional CFO and business advisor, and modelling is most of what the CFO half of that job actually is.
I wrote this because the modelling gap is the single clearest line between a practice that sells advisory and one that talks about selling advisory. Nearly every firm can produce a set of accounts. Far fewer can put a three-way forecast, a downside case and a defensible valuation in front of an owner and then hold their own in the conversation that follows.
The reason isn't capability. It's that the build takes two weeks, and no owner is paying for two weeks.
This playbook is the compressed version of that build.
If you'd rather talk it through than read all of it, book thirty minutes with me.
<aside> ☎️
</aside>
It is one complete advisory model, taken from a set of accounts to a client conversation, with the exact prompt at each stage, a specification of what a correct output looks like before you've seen it, and the review gate that catches the wrong answer.
It isn't a shortcut past your own judgement. Every stage here produces a draft that a qualified person has to check. Where a figure matters, the arithmetic gets rebuilt in Excel. That instruction appears at every stage and it isn't a disclaimer, it's the method.
The business in the worked example is illustrative. Northgate Joinery Ltd doesn't exist. The figures are constructed so they reconcile and so the pattern in them is one you'll recognise, because a made-up example that hangs together teaches better than a real one that has to be redacted until it doesn't. Nothing here is a client case study and nothing here is a claimed result.
Run the first one on a client you know cold. You'll spot a wrong output in seconds. That's the only way to build trust in the pattern, and until you trust the pattern you can't charge for what comes out of it.
Northgate Joinery Ltd. Bespoke commercial joinery, Yorkshire, owner-managed. 18 staff. Year end 31 March. Xero. The practice does year-end accounts, VAT and payroll, and the owner has started asking whether he can afford a bigger workshop.
The worked example runs on Xero because that's where most UK practices are. If the client is on QuickBooks Online, nothing below changes except where the data comes from. The connection section covers both.
Three years of profit and loss:
| FY24 | FY25 | FY26 | |
|---|---|---|---|
| Turnover | 1,850,000 | 2,120,000 | 2,410,000 |
| Cost of sales | 1,184,000 | 1,399,200 | 1,638,800 |
| Gross profit | 666,000 | 720,800 | 771,200 |
| Gross margin | 36.0% | 34.0% | 32.0% |
| Overheads | 540,000 | 596,000 | 651,000 |
| Operating profit | 126,000 | 124,800 | 120,200 |
| Depreciation (in overheads) | 51,000 | 59,000 | 68,000 |
Balance sheet at 31 March 2026: debtors 495,000. Stock and work in progress 270,000. Trade creditors 275,000. Cash 62,000. Bank loan 210,000.