Forecasting revenue from pipeline without wishful thinking

Aug 22
3 min read

Updated: 3 days ago

Introduction

A list of open opportunities and their values is not a forecast. It is a list of hopes with numbers attached.

Turning it into something useful requires two things most small businesses do not have written down: what proportion of deals at each stage historically close, and how long they take.

1. Forecasting revenue from pipeline requires stages with definitions

Stages must be defined by observable events, not by how optimistic anyone feels.

A workable set: enquiry received, qualified as able to buy, proposal sent, verbally agreed, signed. Each transition needs a specific trigger — a proposal has been sent, or it has not.

Stages defined by sentiment produce forecasts that move when moods move. Deals sit in "interested" for months because nobody has to admit anything.

2. Calculate your own conversion rate per stage

This is the number that makes forecasting possible, and it comes from history rather than judgement.

Take the last hundred enquiries, or a year's worth, and count how many reached each stage and how many eventually closed. That gives you a conversion rate from every stage to a sale.

Your rates will differ from any published benchmark, and yours are the only ones that apply. Recalculate annually.

3. Weight each opportunity, then sum

The forecast is the sum of every open opportunity's value multiplied by the historical close rate for its current stage.

A proposal-stage deal worth 10,000 with a 40% historical close rate contributes 4,000. That total is your expected revenue — not the sum of all open deals, which is a fantasy figure.

The weighted number will feel disappointingly small. It is also the one that turns out to be roughly right, which is the entire point.

4. Add timing, or the forecast is unusable

Expected revenue with no date attached cannot inform any decision about hiring, stock or cash.

Use your median time from each stage to close, taken from the same historical sample, to place each weighted amount in a month.

Use the median rather than the average. A handful of deals that took a year will drag an average upward and make your forecast late in a way that is hard to spot.

5. Correct for the biases you will definitely have

Every pipeline forecast is optimistic, in predictable ways.

Deals that have gone quiet are still counted. Stage is recorded as the furthest point ever reached rather than the current reality. Values reflect the largest version discussed. Nobody wants to mark their own opportunity as lost.

Two rules fix most of it: any opportunity with no contact for a defined period drops out automatically, and values record the most likely scope rather than the best case.

6. Do not double count, and watch the pipeline's shape

Two hazards beyond arithmetic.

Do not count both a renewal and a new opportunity from the same client for the same work. And check whether the pipeline is being replenished — a stable total made up of increasingly old deals is a warning, because the forecast holds while the business is actually stalling.

Track average age of open opportunities alongside the total value. Rising age with steady value means nothing is closing.

7. Keep one forecast, and record what it said

Multiple forecasts for the same period, produced differently, mean the business has no forecast.

Publish one, on a fixed schedule, from the same method. Save each version. Then compare what you forecast three months ago against what actually happened.

That comparison is the only way the method improves. Most businesses forecast repeatedly and never check the accuracy of a single one, which is why the same optimism persists for years.

8. Use it for decisions, not for reassurance

The purpose is to make a small number of choices earlier: whether to hire, whether to increase spend, whether to chase work you would otherwise decline.

Set thresholds in advance — if the weighted forecast for next quarter falls below a level, acquisition activity increases. That turns the forecast into a trigger rather than a report.

And keep a low-case version alongside the weighted one, using worse conversion rates. Planning against the low case and being pleasantly surprised is a considerably better way to run a small business than the reverse.

Conclusion

Define stages by observable events, derive close rates and median durations from your own history, then weight every open opportunity and place it in a month.

Age out stale opportunities automatically, record likely rather than best-case values, watch the average age of the pipeline, publish one forecast on a schedule and check it against reality afterwards, and attach decision thresholds so the forecast changes what you do.