Most projections are rejected for the same three reasons. This is what an investor actually tests, and the course that teaches you to build projections that survive it.
Credible projections are built from drivers rather than percentages, reconcile across the profit and loss, cash flow and balance sheet, and hold together when an assumption is changed. Investors do not test whether your forecast is right, because nobody’s is. They test whether the person who built it understands what moves it, and whether the business survives the version where it goes slower.
Having reviewed more than 250 decks and the models behind them, the failures are remarkably consistent. It is almost never a wrong number. It is one of three things.
The growth has nothing underneath it. Revenue rises 40 percent a year because the founder typed 40 percent. There is no customer count, no price, no conversion rate, no capacity constraint — just a curve. An investor’s first question is “what has to happen for this to be true?”, and if the model cannot answer it, the conversation is effectively over.
The costs do not follow the revenue. A hockey-stick revenue line sitting on a nearly flat cost base. Investors know what growth costs: people, marketing, infrastructure, working capital. A plan that triples revenue without tripling anything else reads as someone who has not done the work.
The statements do not agree. Profit is shown, cash is never modelled, and the balance sheet either does not exist or is made to balance with a plug. The moment an investor finds that, every other number on the page becomes suspect.
Three things, in this order.
Can it be traced? Every number should click back to a real input assumption, not sit inside a dense formula that hides four decisions at once. If an analyst cannot audit a line in under a minute, it will be challenged, and an unanswered challenge costs more than a weak number would have.
Does it reconcile? The revenue on the deck, the projections in the model and the figures in the data room must agree line for line. They only do that reliably when the deck is built from the model rather than beside it.
Does it survive the downside? Change the growth rate, lengthen the payment terms, delay the raise by six months. A model that breaks or needs manual patching when one input moves was not built properly. One that shows you exactly what happens to cash is doing its job.
The single most common gap in a founder-built model is that it forecasts profit and ignores cash. They are not the same thing and they diverge exactly when it matters: as the business grows. Receivables stretch, inventory absorbs money, payroll lands before the invoice clears. A business can be profitable on paper and run out of money, and the model is supposed to be the thing that warns you first.
That is why projections that stop at a profit and loss are only a third of the job. The three statements have to be connected, so that a change in revenue flows through to margin, to working capital, and to the cash balance at the bottom of the page.
Course 5 of The CFO Academy takes you from assumptions to a complete, connected set of projections, and then teaches you to read them the way an investor does.
It finishes with an action plan you complete against your own business, not a generic template.
The CFO Academy is six courses sold as one programme with lifetime access. They build on each other, and this one is the foundation the investor-facing documents rest on: the business plan and the pitch deck both draw their numbers from here.
Course 1 is genuinely free and complete, with no card details and no upgrade wall. If you have not worked through it, start there — it is the fastest way to judge whether the teaching is worth your time before you pay for anything.
Financial Modelling & Projections is Course 5 of The CFO Academy: six courses, one payment, lifetime access, from $199. Not ready to pay for anything yet? Course 1 is free and complete.
Explore The CFO Academy