Sandbagging is not a rep problem. It is an equilibrium, arrived at rationally by every participant in the forecasting chain, in which reps under-commit, managers add a buffer on top, and the CRO discounts the whole thing before presenting to the board. Everybody knows this is happening. Nobody can be the first to stop.

A rep with a genuine expectation of closing four deals commits to three, holding one in reserve. This is not deception in the rep's own framing. It is prudence, born of the certain knowledge that a missed commit is professionally expensive and a beaten commit is professionally free. The rep is buying insurance with a currency the organisation has told them is cheap.

The frontline manager, aware that reps under-commit, applies a mental correction upward. The CRO, aware that managers correct upward but uncertain by how much, applies a further adjustment. Each layer is compensating for a distortion introduced by the layer beneath it, and each layer's compensation introduces a new distortion of its own. The number that reaches the board has passed through three deliberate adjustments and bears an unknown, unstable relationship to what the sales organisation actually expects to happen.

The equilibrium is stable because unilateral defection is punished. A rep who commits honestly and misses is worse off than a rep who sandbags and beats. A manager who stops adding buffer will look worse than peers who continue. The behaviour persists not because anyone endorses it but because nobody can move first.

"
Everybody knows this is happening. Nobody can be the first to stop.

Calculate the ratio of actual closed revenue to committed revenue, by rep, over eight or more quarters. In a healthy forecasting culture this ratio should cluster tightly around 1.0, with roughly symmetrical misses above and below. In a sandbagging equilibrium the distribution is asymmetric: reps consistently beat their commit by ten to thirty per cent, and almost never miss.

That asymmetry is the diagnostic. Genuine forecast uncertainty produces misses in both directions. A team that beats its commit in seven quarters out of eight is not forecasting. It is negotiating a floor and then clearing it.

Cross-reference against the timing of the beat. Reps who sandbag typically close their held-back deals in the final week of the quarter, having deliberately parked them. A concentration of unexpected closings in the last five days, from reps who had not flagged those deals as likely, confirms the mechanism.

The asymmetry of consequence is the whole explanation. Missing a commit generates a difficult conversation, a mark against the rep's credibility, and in some organisations a performance conversation. Beating a commit generates a brief moment of praise and no lasting benefit. Faced with an asymmetric payoff, reps rationally position their commit below their expectation, and the size of the buffer scales with the severity of the penalty for missing.

Managers are not naive about this. They know the number they receive is conservative and they adjust for it. But because the size of each rep's buffer is unknown and varies with the rep's confidence and risk tolerance, the manager's correction is a guess. The whole system converges on a forecast that everybody privately distrusts and publicly defends.

A sandbagged forecast is not simply pessimistic. It is uninformative. The board receives a number that has been adjusted three times by parties with incomplete knowledge of each other's adjustments, and the confidence interval around it is unknowable from the outside. Historical forecast accuracy in such a system tells an investor nothing about forecasting capability, because the accuracy was manufactured by setting a low bar rather than by predicting well.

The practical consequence in a transaction is that the forward revenue model cannot be calibrated against the target's own forecast, because the forecast is a negotiating position rather than an estimate. Any acquirer who assumes a sandbagged organisation will continue to beat its numbers post-acquisition is assuming that the equilibrium survives a change in ownership, incentive structure, and management, which it frequently does not.

Risk Classification: Behavioural Risk (primary) / Forecast Risk (secondary)
Behaviour Observed
Reps systematically commit below their genuine expectation, managers apply an upward correction, and senior leadership applies a further adjustment, producing a board-level forecast that has been distorted three times by parties with incomplete knowledge of each other's adjustments.
Why This Happens
Missing a commit is professionally expensive and beating one is professionally free. Faced with an asymmetric payoff, reps rationally position their commit below expectation, and unilateral defection from the equilibrium is punished.
Investment Risk
Historical forecast accuracy measures the height of the bar rather than the quality of the prediction. The forward model cannot be calibrated against the target's forecast because the forecast is a negotiating position, and the equilibrium producing it rarely survives a change of ownership.
Implication for the Investment Committee
Calculate the distribution of actual-to-committed revenue by rep across eight quarters. Asymmetric overperformance with almost no misses indicates a sandbagged forecast whose accuracy carries no predictive information.
Valuation Risk MEDIUM
Forecast Risk HIGH
Execution Risk LOW