Skip to content
Enjoy complimentary data migration when switching from your existing bookkeeper or CPA to Institution.

Bookkeeping & finance

Revenue forecasting: building a forecast from real drivers

Build revenue assumptions from customers, pipeline, prices, renewals, units, or capacity, then compare the forecast with actual results.

By Institution3 minute readUpdated View sources
In this guide

Start with what causes revenue

A growth-percentage forecast provides a useful reference, but the operating plan also needs a reason that revenue can reach that level. Identify the limiting inputs: delivery capacity, demand, conversion, pricing, and the contract conditions that determine when revenue is earned.

For a services firm, start with available people, billable hours, realized billing rates, and delivery timing. For a store, examine traffic, conversion, order size, and returns. For a subscription business, separate new sales, renewals, expansion, and churn. Map bookings and billing to the revenue-recognition policy before using them as recognized revenue. Customer collection timing belongs in the connected cash forecast.

Six consultants: $88,200 revenue, separate receipts

Suppose six consultants each have 150 available hours for the month, 70 percent are billable, and the realized rate is $140. If the work is performed, billable, and recognizable during that month, modeled revenue is 6 × 150 × 0.70 × $140 = $88,200. At 65 percent utilization it is $81,900, a $6,300 reduction. Add a seventh consultant using their actual planned start and ramp assumptions.

Now model receipts separately. If 95 percent of those same-month invoices are expected to be collected in the month, they contribute $83,790 of receipts in the first case. The remaining $4,410 stays in the collection schedule, subject to any documented credit-loss assessment. Receipts from prior months' invoices must be added separately. Multiplying revenue by a collection percentage does not convert that cash figure into accrual revenue.

References: SEC: Beginners' guide to financial statements

Check it against history

Run your model on a past period and compare it with what happened. If the model would have been 15 percent off last year, adjust it before you rely on it. Use actual results from closed books for the historical base.

Keep actuals and forecast apart

Show closed months as actuals and open months as forecast, and mark the difference. A model that overwrites history with projections makes it impossible to judge the forecast later.

Update and explain

Revise the forecast each month, and note what changed and why. Keep the number of drivers small. A model with a few well-understood drivers is easier to trust than one with fifty guesses.

Three methods, and when each fits

A bottom-up forecast builds from drivers, as in the example above. It is the best choice when you can name what causes sales. A trend or run-rate forecast extends recent results, adjusting for seasonality. It works for stable businesses and as a sanity check on the bottom-up view. A top-down forecast starts from the size of the market and assumes a share. It is common in investor materials but weak as an operating tool, because it does not say how you will win those sales.

Use at least two and compare. A large gap between them is a question worth answering.

Mistakes to avoid

Assuming a good month will repeat. Ignoring seasonality. Forecasting revenue without the matching cash timing. Hiding uncertainty by giving one number instead of a range. Consider a base case, a downside, and an upside, and plan spending against the downside.

Frequently asked questions

How far ahead should I forecast?

Match the horizon to the decision. A twelve-month operating view and a thirteen-week cash forecast are useful starting points; long leases, capital projects, or financing commitments may need a longer view with wider uncertainty ranges.

How often should I update it?

Monthly, after the books close. Update sooner after a major win or loss.

How accurate should it be?

Track error by revenue stream and forecast horizon against the decisions the model supports. A small total error can hide offsetting misses. Investigate repeated bias and judge the model against a simple historical baseline rather than adopting an unsupported universal percentage target.

Sources

Requirements and product details can change. Check the linked guidance for the facts that apply to your situation.

Related reading

Need help with this work?

General information, not personalized accounting, tax, legal, or investment advice. Have the appropriate qualified professional review decisions that depend on your circumstances.