Two numbers, two accountabilities
A rep forecast is what a seller commits to for their own deals, and a manager forecast is what their leader commits to for the team. The two are separate submissions, not one number passed upward. The company runs on the manager figure, but keeping the rep figure intact is what makes either of them gradeable.The manager number exists because reps see one territory and managers see patterns across several. A seller cannot tell that their commit deal looks exactly like six deals that fell out last quarter. Their manager can.
The spread is the measurement
Store both numbers and track the difference each period. That spread is the clearest available record of a manager's forecasting judgment.
- A consistently negative spread means the manager cuts the roll-up. If their number lands closer to actuals than the roll-up did, the cut is skill. If it does not, the manager is applying a reflex. - A consistently positive spread means the manager adds to what their team called, usually on deals they are personally working. - A near zero spread every period means no judgment is being applied and the manager level is a pass-through.
Grade both numbers against the actual result at the end of each period. Over four quarters this separates managers who improve the forecast from managers who add noise, and it does so with evidence rather than impression.
Adjust deals, not totals
A percentage haircut applied to a team total cannot be verified later. A deal level adjustment can. When a manager cuts a rep's commit deal from 200,000 to zero, that decision gets graded when the deal closes or does not.
ORM identifies a rep changing the close date as the strongest signal of deal slippage, and finds that a deal that has already slipped once becomes less likely to close even while it sits in commit. That makes close date history the natural first filter for a manager deciding which rep calls to discount.
The reverse pattern matters too. A rep whose commit consistently converts above the team rate is likely holding deals back, and cutting their number compounds a distortion that already exists.
Make the adjustment visible
Publish the adjustments with the submission. Reps who learn that their number gets trimmed by a fixed amount will pad the next one, and within two quarters neither figure means anything.
For how the levels stack and what each is allowed to change, pair this with forecast accuracy and how to create a sales forecast.
Frequently Asked Questions
Should a manager forecast equal the sum of their reps?
Rarely. A manager who always submits the exact roll-up is adding no judgment and cannot be graded separately. A manager whose number never resembles the roll-up is running a parallel forecast and should explain which deals they are discounting.
How much should a manager adjust the roll-up?
Enough to reflect deals they have inspected and disagree with, and no more. Adjustments should be deal specific rather than a percentage haircut applied to the whole team, because a haircut cannot be checked against anything later.
Should reps see the manager number?
Yes, along with which of their deals were adjusted and why. Cutting a rep's number silently teaches the rep to inflate the next submission to survive the expected cut.
Which number should the company use?
The manager number rolls up, since that is the level accountable for the team result. Both numbers should still be stored and graded so you know whose judgment improved the forecast and whose degraded it.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like rep forecast vs manager forecast into prescriptive action for your team.
Schedule a Demo