Blog7 min

How to forecast revenue when market conditions change

When conditions shift, the first thing to break is any forecast derived from historical conversion rates, because it assumes this quarter's deals behave like last year's. Stop leaning on those rates, re-score each open deal on evidence rather than stage, widen the range you report, and watch slippage rather than win rate — slippage moves months earlier.

What breaks, and in what order

Cycle length goes first. Deals do not start failing, they start taking longer. An extra approver appears, a decision moves to the next budget cycle. Your win rate looks unchanged for a quarter or two because the deals have not resolved yet.

Then slippage rises. The share of deals forecast for a month that actually closed in it starts falling. This is the earliest number that moves, which is why it belongs on your dashboard before any of the obvious ones.

Win rate goes last. By the time it has fallen visibly, the deals concerned entered the pipeline two cycles ago. It is a lagging indicator and a poor early warning.

A team watching win rate in a downturn finds out roughly two quarters late. A team watching slippage finds out in the first month, because deals start moving before they start dying.

Four adjustments that work

1. Re-score every open deal on evidence, not stage. For each one, ask what the buyer has actually done that costs them something: a technical review, legal engaged, a named budget line. In a tightening market, verbal enthusiasm decouples from outcome far more than usual, while costly signals keep their meaning.

2. Add an explicit budget question. Not "is there budget" — the answer is always yes — but "has this been approved for this fiscal period, by whom, and what would cause it to be revisited". Deals die in reforecasts, not in meetings.

3. Widen the range you report. If your inputs are less reliable, your output is less certain, and saying so is more credible than a confident number that gets revised twice. A range that widened and was honest about why survives a bad quarter; a point estimate does not.

4. Shorten your forecast horizon. Quarterly forecasts built three months out are nearly worthless when conditions move. Forecast the current month with confidence and treat the quarter as a range.

What not to do

  • Do not apply a blanket haircut. Taking 20% off everything feels prudent and destroys information: it treats the deal with a signed security review identically to the one that has gone quiet.
  • Do not switch methods mid-downturn. You lose comparability exactly when you most need to know whether things are getting worse. Keep the method, change the inputs.
  • Do not discard your history. Your bias — how much you habitually over- or under-forecast — tends to persist through conditions even when conversion rates do not. It remains the most reliable correction you have.

When conditions improve

The same logic inverts, and teams are much slower to notice. Cycles shorten, deals that were parked come back, and a pipeline scored during the downturn is now systematically pessimistic. The tell is the same metric: slippage falls, and deals start closing earlier than forecast.

If you kept the record through the bad quarters, you will see the turn in your own numbers before you see it in anyone's market commentary. That is the compounding argument for keeping it: not that it makes any single forecast right, but that it is the only thing that tells you when your assumptions have stopped being true.

All articles