When a rep consistently closes 120–130% of their stated commit quarter after quarter, there are two ways to interpret it. The first is that they're an exceptional closer who habitually finds upside in the last week of the quarter. The second is that they're systematically under-committing — holding deals back from their forecast submission to protect their number, then releasing them as "surprise" closes. The second interpretation is sandbagging, and it's a forecasting problem regardless of how it reads on the commission statement.
Sandbagging is rarely malicious. It's a rational response to incentive structures that punish missing more than they reward exceeding. A rep who committed $800K and closed $720K got a difficult conversation. A rep who committed $600K and closed $720K looks like a hero. If the incentive gradient is asymmetric, you'll get sandbagging — not because your reps are bad actors but because they're logical ones.
The problem for sales leaders is that sandbagging makes your forecast systematically wrong in the conservative direction, and it's invisible to most standard forecast analysis because the rep consistently looks like they're over-performing. You don't see the commits that should have been higher.
What Sandbagging Looks Like in the Data
Sandbagging has a specific statistical signature. Across a rep's history of four or more quarters:
- Their commit-vs-close ratio is persistently above 1.0 — they regularly close more than they committed, by a margin that exceeds random variance (10–15% would be noise; 25–40% quarter over quarter is a pattern).
- Their late-quarter "surprise" closes follow a predictable pattern. Deals that weren't in their committed forecast appear in the final week of the quarter, often already past proposal stage and close to signing. These weren't truly surprises — they were in a closeable state earlier but kept in best-case or below.
- Their best-case number is substantially higher than their commit — more than 50% higher — and their actual close lands in or near the best-case range rather than the commit range.
Each of these individually could have non-sandbagging explanations. Together, over four or more quarters, they're a strong indicator of systematic under-commitment.
The Activity Pattern Dimension
Where activity data adds something that pure commit-vs-actual analysis doesn't: you can see when a deal was behaviorally closeable versus when the rep committed it. A deal that a rep first included in their commit in the final week of a quarter, but where the activity data shows active multi-threading, executive-level engagement, and a fast email response cadence for the prior six weeks — that's a deal that was in a close-ready state well before the rep committed it.
The divergence between "when the deal was behaviorally ready" and "when the rep included it in forecast" is the sandbagging gap. It's measurable if you have deal-level activity data and historical commit records.
Consider a scenario at a growing enterprise software vendor. Their VP Sales noticed that one of their top-performing reps had closed at 128%, 134%, and 119% of commit over three consecutive quarters. Looking at the activity data for the deals that appeared in the final weeks of each quarter, those deals had active engagement patterns — meeting attendance, multi-stakeholder email threads, procurement contact established — dating back four to six weeks before they entered commit. The behavioral readiness signal was there; the rep just hadn't surfaced it in the forecast.
Why This Is a Forecasting Problem, Not Just a Trust Problem
A VP Sales or CRO who relies on sandbagging reps to provide upside surprise will systematically under-forecast to their own board. If you know three of your reps habitually close 25–30% above commit, you'll build a mental adjustment into your board number — but that adjustment is informal, undocumented, and not explainable if challenged. It also creates a dependency: if one of those reps has a genuinely tough quarter, their upside doesn't arrive, and your implicit assumption turns into a miss.
The deeper issue is that sandbagging corrodes the forecasting methodology. If commits are widely understood to be conservative floors rather than genuine probabilistic assessments, the commit number stops being useful information. Pipeline reviews devolve into everyone knowing that the real number is somewhere between commit and best-case, and decisions get made on vibes rather than evidence.
This Is Not an Argument That All Upside Is Sandbagging
This is not to say that a rep who occasionally closes above commit is sandbagging. Quarter-end acceleration is a real phenomenon — buyers who have been deliberating sometimes sign when they perceive urgency. Good reps sometimes work deals that genuinely weren't closeable until the last week. One or two quarters of over-commit is noise. A consistent four-quarter pattern with the specific activity signature described above is worth examining.
There's also a distinction between a rep who is legitimately cautious about calling deals that aren't certain and one who is deliberately holding back commitments to manage their number. The cautious rep typically has deals where both activity and commit are late to materialize. The sandbagger has deals where activity shows readiness weeks before the commit does. That's the key diagnostic difference.
Having the Conversation Without Making It Adversarial
If you've identified a sandbagging pattern through data, the conversation with the rep should lead with the data observation, not an accusation. "I've been looking at your last four quarters, and I notice that you've closed three or four deals per quarter that weren't in your committed forecast — and the activity data shows they had strong engagement profiles four to six weeks before close. Help me understand how you were thinking about those deals when you submitted your commit." That's a fact-based question, not an accusation.
The rep's answer will be informative either way. They might explain a legitimate reason — deals that genuinely accelerated, deals where they had real uncertainty about the timeline. Or they might confirm that they hold back some deals deliberately. Either answer helps you understand what your forecast numbers actually mean.
Structural Fixes: Incentives and Forecast Discipline
The behavioral fix for sandbagging is changing the incentive gradient. If over-committing is punished but under-committing is implicitly rewarded, you'll get conservative commits. Two approaches that address this directly: first, include forecast accuracy (commit vs. actual delta within a band, say ±10–15%) as part of the annual performance review — it makes accuracy valuable, not just results. Second, introduce a "best-case" commit category that reps are encouraged to populate, making it safe to surface high-probability deals without the full commitment risk of the primary commit column.
Activity data helps on the detection side: when you can show a rep that their behavioral pattern points to a deal being in close range, and that this observation is going into your planning regardless of what they commit, the strategic value of holding the deal back decreases. If the VP already knows the deal is probably closing because the scoring model says so, the rep's commit number is less of a protective shield.
The goal is not to eliminate upside — genuine Q4 acceleration and late-breaking deals are valuable. The goal is to eliminate the category of deals that were predictably closeable weeks before quarter end but were held back from the forecast, because those deals are systematically distorting your ability to produce an accurate plan.