Sales Forecast

Most of a company's spending decisions look several months ahead. Whether to hire three engineers, double the ad budget, or move offices all depend on money that hasn't arrived yet. In a company whose revenue comes from contracts, that money is not only unreceived — which deals will close isn't settled either.

So the first meeting of every quarter ran the same way. The CEO would ask how much was coming in, and the sales lead would answer that it should reach $230,000 if things went well. What was packed inside "if things went well" nobody knew. Hire against that number, finish the quarter at $138,000, and hiring freezes the next quarter — which drops the revenue after that as well.

A sales forecast turns that one sentence into a procedure. Each open deal gets laid out with its stage, its value and its expected close date, and historical close rates get applied and summed. The result is still an estimate, but it's an estimate with a traceable origin. That's the difference from a gut number: when it's wrong, you can find out where it went wrong. If the forecast said $230,000 and $138,000 landed, you can point at which stage's close rate was set too high, or which deal slipped and why, and correct the next one.

The same six steps repeat each quarter.

1. Clean the deal list. Keep only deals with an expected close date inside the period, and bring the value and the deal stage up to date. Deal stage is the field that records how far along a deal has come.
2. Attach probability. Multiplying by per-stage close rates is the most common method. Ten deals worth $15,000 each at the proposal stage, and a historical close rate of 40 percent for that stage, gives an expected $60,000.
3. Have reps add a call. Separately from the arithmetic, each deal goes into one of three buckets: near certain, possible, unknown. Facts like whether the signer's approval chain has been confirmed aren't in the data, and only the rep has them.
4. Have the lead adjust. Apply each rep's accuracy over the last four quarters. A rep who has historically delivered 70 percent of what they called turns a $230,000 submission into about $161,000.
5. Submit a single number. What the company plans against has to be one figure, not a range. Publish a range and each recipient picks a different point inside it.
6. Compare after the quarter closes. Put the forecast and the actual side by side and write down what explains the gap. That record is the raw material for step 4 next quarter.

Step 6 is the one that gets skipped. Without the comparison there's no accuracy figure, and the adjustment in step 4 goes back to being a feeling.

Reps submit first. The person who knows a deal best drafts the number, and the sales lead consolidates and adjusts. Sales operations — the role that runs the tooling and data the sales team depends on — aggregates the figures and calculates the per-stage close rates. Finance receives the result last and folds it into the spending plan.

The timing splits three ways. A full forecast gets built in the first week of the quarter. It then gets refreshed every two weeks, and that cadence only works if it is shorter than the sales cycle. With deals averaging 34 days, reviewing the forecast once a quarter means finding out about a slip after the quarter has ended. The third moment is the week after close, when forecast meets actual.

The final three weeks are where it moves most. Deals that arrive in that stretch can't realistically close inside the period, so they're no basis for raising the number, while a deal treated as certain that stalls in the customer's approval chain removes $23,000 in one move. So in those last three weeks the meeting shifts away from counting new deals and toward confirming signature logistics on the deals already there.

"I'm submitting $185,000 for the quarter — $123,000 commit, $62,000 in the possible column."

Reporting the certain and the uncertain as separate figures rather than one blended one.

"Legal review hasn't finished on that one. Why is it in commit?"

Checking the basis for a call, which is how inflation gets caught.

"Last quarter's forecast came in 30 percent above actual, so I'm discounting this one by the same amount."

Adjusting a submitted number using the rep's historical accuracy.

At Pagedeck this meeting runs every other Monday. Sofia reads through her deals and Grace asks for the basis behind every item sitting in commit. The questions are fixed at three: has the signer been confirmed, how many days does the customer's internal process take, and when was the last contact. Any one of those blank and the deal comes out of commit.

When the quarter closes, Grace writes the forecast and the actual next to each other. A rep who has come in high two quarters running doesn't get a lower target — the commit standard gets recalibrated instead. Comparison has to be against revenue closed won, and a deal whose signature slid to the first day of the next quarter is not this quarter's result.

The most common mistake is treating the forecast as the target. Both are future revenue figures discussed in the same meeting, so they blur together easily. But a target is a number set in order to be reached, and a forecast is a number worked out to describe what the current state produces. Treat them as one thing and the forecast drifts upward toward the target: reps find it uncomfortable to submit below the target, and the company raises spending on the inflated figure. At quarter end, the shortfall and the money already spent arrive together.

The other mistake is treating forecasting as a game to be won. Accuracy gets tied to evaluation, so the response is predictable. Once hitting the number becomes the goal, reps call low and overshoot. That behavior has a name — sandbagging — and it looks like a rep who has already cleared the quarter pushing a contract signed in the final week into the first week of the next month. Their personal accuracy improves while the company plans two consecutive quarters against the wrong figure. Which is why accuracy belongs in the adjustment coefficient rather than in a performance review.