A forecast is a decision tool

Forecasts usually exist because someone outside the business asked for one: a bank, a board, an investor or a grant application. A spreadsheet written for a reader is written to pass. Revenue climbs at a confident angle, costs stay polite and cash never quite runs out. Once it is sent, nobody opens it again.

The commercial cost is not the hours spent building it. It is the decisions made without it. Nobody knows what revenue will be in month fourteen, including the person who typed it in. What a forecast can do is put every assumption where you can see it, show which few actually move the result, and tell you what you would have to believe for a particular decision to be a good one.

Write the decision at the top of the sheet

Before building anything, write the decision in one line. Different decisions need different models, and a model that tries to answer everything answers nothing well.

• "Should we hire a second salesperson in Q2?" needs ramp-up time, conversion rate and contribution margin per deal.
• "Can we open a second location without new borrowing?" needs fit-out cost, months to break-even and the lowest cash point in between.
• "How long can we run before we need outside capital?" needs a month-by-month cash view. An annual P&L will hide the answer.

A model built for one question is smaller, quicker to check and easier to trust. It also has an end date. Once the decision is made, the model has done its job and can be retired or turned into a monthly tracker.

Build scenarios that describe different situations

The usual low, base and high set is often one story told at three volumes, with every line scaled by the same percentage. That teaches you nothing. Each scenario should change a specific assumption for a specific reason.

• Cautious: uncomfortable but plausible. The sales cycle takes twice as long as you hope, or a key customer pays at 60 days instead of 30.
• Base: anchored to what you have observed, such as last year's figures, pilot results or signed letters of intent.
• Ambitious: names the one or two things that must go right, such as a price increase holding or a conversion rate improving.

The scenarios are not there so you can pick the correct one. They are there so you decide in advance what you would do in each, and so you can see early which one you are drifting towards.

Find the two assumptions that decide the outcome

In most small-company models, a handful of inputs drive nearly everything: price, volume, gross margin, payroll and payment terms. Move each one by ten or twenty percent and watch the cash line. Usually one or two of them dominate. Those deserve your validation effort and your management attention.

A long debate about whether a software subscription costs 4,200 or 4,600 SEK a month is a sign the model is being used to feel busy. If a number cannot change the decision, round it and move on.

Keep an assumptions register

Every number that is not a recorded fact belongs on one visible page, with its source and how confident you are in it. Two illustrative entries:

• "Average order value 1,850 SEK. Based on 64 orders, March to May."
• "Monthly churn 3%. Rule of thumb, not yet observed in our own data."

A lender or investor reading this page can see that you know the difference between what you have measured and what you are hoping for. That does more for credibility than formatting ever will.

Compare forecast to actual every month

The value builds when the forecast meets reality. Each month, ask where the gap came from: price or volume, timing or demand, one customer or many. Each answer improves the next version. After six months you have a model that describes how your business actually behaves, which very few small companies have.

Checklist before you rely on a forecast

• What decision is this for, and by what date must it be made?
• Which few assumptions drive most of the outcome?
• What would we do differently in the cautious case?
• In which month does cash reach its lowest point, and how low?
• When will we next compare this to actual figures?

A forecast that changes what you do is worth more than one that impresses whoever reads it.