02 — Plan
Nobody funds a company that cannot forecast itself.
Clean books tell an investor you are competent. A three-way model that ties to those books, holds under scenario, and has been right before tells them you can be trusted with their money. This is the step between running the finance function and raising against it — and it is the one most companies skip.
Where this sits
Run, then Plan, then Raise.
The sequence is not arbitrary. Each stage is the input to the next, and skipping one is what makes the next one expensive.
Clean, current actuals
Five-day close, substantiated balance sheet, compliance filed. Without this, everything downstream is fiction with a spreadsheet attached.
A model that ties
Three integrated statements, driven by operating variables you control, reconciling to the actuals every month. This page.
Capital on your terms
Diligence tests the model harder than the deck. Preparation here is what converts into valuation there.
Oversight that holds
A board that can read the forecast, challenge the assumptions and hold management to the plan.
Three-way forecasting
Profit is an opinion. Cash is a fact. A three-way model reconciles the two.
Most founder-built models are a profit and loss with a cash line bolted underneath. They cannot tell you what happens to your bank balance when debtor days slip by a week, or why a profitable quarter still consumed capital. An integrated model can, because the three statements are wired to each other rather than typed independently.
Build standard
Eight rules we will not break.
A model that only its author can operate is a liability. These are the conventions that make ours transferable, auditable and survivable.
Inputs are separate and coloured
Every assumption lives on one input sheet, formatted so it is unmistakable. No hardcoded numbers buried inside a formula, anywhere, ever.
One formula per row
Copy right across the whole timeline without exception. The moment a row has two different formulas in it, the model has a bug you have not found yet.
Actuals and forecast in one timeline
Closed periods carry actuals from the ledger; open periods carry forecast. Variance is then a formula, not a monthly rebuild.
Checks that fail loudly
Balance sheet balances, cash ties to the cash flow, opening equals prior closing. A visible check row that turns red beats a quiet error every time.
Drivers, not growth rates
"Revenue grows 8 per cent" is a wish. "Sales hires × ramp × quota × win rate" is a plan you can manage against and an investor can interrogate.
Scenarios by switch
Base, downside and upside driven by one selector, not three saved copies of the file that quietly diverge after the second week.
Every assumption sourced
A note against each input saying where the number came from and who owns it. This is the first thing diligence asks and the last thing anyone documents.
Accuracy tracked against itself
We report how wrong the last forecast was, by line. It is uncomfortable, and it is the only way anyone learns how much to trust the next one.
Group & consolidation
One entity is arithmetic. A group is a discipline.
The moment you have a second entity — an Australian subsidiary, a property holding company, an employee share trust — forecasting stops being a bigger spreadsheet and becomes a structural problem. Most groups discover this during a raise, which is the worst possible time.
Entity-level, then consolidated
Each entity forecast in its own functional currency with its own drivers, then consolidated with eliminations — not a single blended model that no subsidiary board could ever sign.
Intercompany that eliminates
Management charges, loans, interest and transfers modelled on both sides so they cancel. If your consolidated revenue includes an intercompany sale, the diligence adjustment will be brutal.
FX handled properly
Average rate for profit and loss, closing rate for the balance sheet, translation differences to reserve. Forecast at policy rates with a sensitivity, not at whatever the rate was on the day you built it.
Cash where it actually is
Group cash is not one number. It is cash by entity, by currency, net of what is trapped by tax, thin capitalisation, minority interests or a lender's cash sweep.
Covenants at the tested level
Facilities test at a specified entity or sub-group. The forecast has to produce that exact perimeter, or your headroom number is decorative.
Segment view for the board
Directors govern by division and geography, not by legal entity. The same model has to slice both ways without a second build.
Trapped cash is the most common unpleasant surprise in a group forecast. The consolidated balance shows twelve million; the entity that needs to make payroll on Thursday holds four hundred thousand.
The planning rhythm
Forecasting is a cadence, not a document.
13-week direct cash
Receipt by receipt, payment by payment, built directly rather than derived from profit. It is the only forecast granular enough to make an operating decision on — whether to take the discount, delay the hire, or call the facility. Rolled forward every Monday with last week's actual against last week's forecast shown side by side.
Direct method · By entity and bank account · Actual vs forecast varianceRolling 12-month cash
The horizon a board and a bank actually govern to. Twelve months forward, rolled one month every month so it never shortens, built from the same drivers as the three-way model rather than maintained separately. The 13-week view answers "can we pay for this"; the 12-month view answers "can we commit to this" — and they must reconcile to each other, which is the discipline most companies skip.
Rolling horizon · Reconciles to the 13-week · Base, downside and upsideRe-forecast and variance
Actuals load, the model re-forecasts, and we write the commentary: what moved, whether it is timing or permanence, and what it does to the full-year landing point. The question is never "did we hit budget" — it is "what does this month tell us about the rest of the year".
Full-year reforecast · Variance bridge · Landing pointScenario and stress
Base, downside and upside re-run against the drivers that have actually moved. Where is the break point on runway, on covenant headroom, on gross margin? A board that has seen the downside modelled before it arrives makes calmer decisions when it does.
Scenario switch · Sensitivity table · Break-point analysisBudget and plan build
Bottom-up from the operating teams, reconciled to the top-down expectation the board and investors hold, with the gap made explicit rather than averaged away. Signed off, locked, and then used as the comparator all year rather than quietly revised.
Bottom-up build · Top-down reconciliation · Board approvalBusiness partnering
Finance in the room where the decision is made.
Reporting tells the business what happened. Partnering changes what happens next. It means a finance person sitting with the sales lead on pricing, with the operations lead on capacity, with the founder on the hiring plan — before the commitment, not in the variance report afterwards.
Pricing & margin
Contribution by product, channel and customer. Which revenue is actually worth having, what a discount really costs, and where price can move without volume following.
Unit economics
Acquisition cost, payback period, cohort retention and lifetime value — defined once, consistently, so the number does not change depending on who is presenting.
Operational modelling
Utilisation, throughput, headcount to volume, and the point at which the next fixed cost step has to be taken. Where finance and operations actually meet.
Business cases
A consistent appraisal standard for every material spend — payback, net present value, sensitivity and the honest downside — so proposals compete on merit rather than advocacy.
Cash conversion
Debtor days, creditor terms, inventory turns and the cash release available inside the business before anyone raises external capital.
Incentive design
Targets and commission structures that pay for the behaviour you want and are affordable in the downside case as well as the base.
FinOps & transformation
The forecast is only as good as the operation feeding it.
If quoting lives in one system, delivery in another and invoicing in a spreadsheet, no model will ever be accurate — because the data arrives late, incomplete and manually re-keyed. Fixing the forecast usually means fixing the workflow underneath it.
Map the flow
Quote to cash, procure to pay, hire to retire. Every handoff, every system, every re-keying, every approval that waits on one person's inbox. We time each step and cost the delay. Most businesses have never seen this drawn, and the first look is usually uncomfortable.
Process map · Cycle time · Cost of delayFix the data at source
Automation on top of bad data produces wrong answers faster. Chart of accounts restructured, tracking dimensions defined, master data cleaned, and the system of record agreed for each fact — one place per number, no exceptions.
XL Standard chart of accounts · Dimensional model · Single source of truthAutomate the mechanical
Invoice capture and coding, approval routing, bank reconciliation, payment runs, expense handling, revenue recognition schedules, intercompany journals. Supervised agents do the volume; a Chartered Accountant reviews the exceptions.
Xero · ApprovalMax · Hubdoc & Dext · payment rails · XLCFO OS agentsWire operations to the model
Pipeline from the CRM, usage from the product, hours from the delivery system, headcount from payroll — feeding the drivers directly. When operational reality changes on Tuesday, the forecast knows on Tuesday, not at the next month-end.
Live driver feeds · Automated reforecast · Exception alertsRedesign the team around it
Once the mechanical work is gone, the finance team's job changes and the structure has to change with it. Fewer processors, more analysts and partners. We rewrite the roles, the responsibilities and the calendar — and say plainly which positions the new operating model no longer needs.
Operating model · Role design · Capability planWhat you get
Deliverables, not a workshop.
| Deliverable | What it is | Cadence | Who uses it |
|---|---|---|---|
| 13-week cash forecast | Direct-method receipts and payments by entity and account, with prior-week variance | Weekly | Founder, CFO, treasury |
| Rolling 12-month cash forecast | Twelve months forward, rolled monthly, reconciling to the 13-week view | Monthly | Board, lenders, founder |
| Three-way model | Integrated P&L, balance sheet and cash flow, driver-based, scenario-switched | Monthly refresh | Board, investors, lenders |
| Consolidated group forecast | Entity-level models with eliminations, FX and covenant perimeter | Monthly | Group board, lenders |
| Variance bridge | Budget to actual to reforecast, decomposed by driver rather than by account | Monthly | Management, board |
| Scenario pack | Base, downside, upside with break points on runway and covenants | Quarterly | Board, audit & risk |
| Annual budget | Bottom-up build reconciled to top-down expectation, board approved | Annual | Whole business |
| Business case template | House standard for appraising material spend consistently | On demand | Department heads |
| FinOps roadmap | Process map, automation plan, sequenced with owners and expected benefit | Project | Executive team |
Included in the Control tier and above. Standalone model build and FinOps transformation are also available as fixed-price projects — see pricing.
Send us your current model.
We will stress-test it and come back with what breaks, what an investor will challenge, and what we would rebuild. No charge for the review.