Sandbagging Is a Comp Plan Problem, Not a Discipline Problem
You can coach a rep on forecast honesty for a year and never beat what the accelerator curve is quietly telling them to do.
Every sales VP has had this conversation: a rep sits at 95% of quota with twelve days left in the quarter, holding two deals that are ready to sign. They close one. The other slides into next quarter, three days after the reset, even though the paperwork was ready last Thursday. The VP calls it "a discipline issue" and puts the rep on a coaching plan for forecast honesty. The rep nods, says the right things, and does exactly the same thing next quarter. Not because they're dishonest. Because the comp plan is telling them, in numbers, that holding the deal is the correct decision — and no amount of coaching beats arithmetic.
I've run comp plan reviews for three companies as CCO and audited a dozen more as a consultant, and the pattern is consistent enough that I'll say it plainly: sandbagging is not a trust problem. It's what happens when a rational person reads an accelerator curve, a true-up window, or a ratchet clause correctly. If you want less of it, redesign the plan. Coaching the rep is coaching the wrong layer of the system.
The three mechanics that actually cause it
Sandbagging isn't one behaviour with one cause. In every plan I've pulled apart, it traces back to one — usually two — of three specific structural features.
1. The accelerator cliff. Most plans pay 1x commission up to 100% of quota, then jump to 1.5x or 2x above it, sometimes with another step at 120% or 150%. A rep sitting at 92% with a deal that would take them to 101% has every reason to close it now: it unlocks a materially higher rate on everything after. But a rep already at 140%, deciding whether to close deal 41 this week or next, is doing different maths. If the accelerator resets to zero next quarter, closing now versus in eleven days makes no difference to the rate — so there's no cost to timing it for whichever period needs the boost more. The cliff doesn't create pressure to under-report; it creates pressure to time reporting, and a forecast built on gut-feel percentages can't tell the difference between the two.
2. The ratchet. This is the one managers underrate. If next year's quota is set as a multiple of this year's actual attainment — common in growth-stage SaaS — then a rep who closes 160% this year is handed a materially harder number next year for the same OTE. That rep isn't being difficult when they hold three deals until January. They've read the spreadsheet correctly. Punishing over-performance with a bigger number is a subscription-model tax on honesty, and reps figure this out by their second year even if nobody ever explains it to them directly.
3. The true-up and decelerator. Some plans cap payout above a ceiling, or step the rate down past a certain point to guard against "windfall" commission. Once a rep clocks that deal 43 pays less per pound than deal 22, the incentive to report it this period evaporates. Add a true-up period at year-end that lets held deals count retroactively, and you've built a formal, sanctioned mechanism for exactly the behaviour you're disciplining people for.
What this looks like on a forecast call
None of this shows up as "I am sandbagging." It shows up as:
| What the rep says | What's actually happening |
|---|---|
| "Just tightening up the paperwork before I move it to commit" | Deal is signed-ready; moving it now doesn't help this period's accelerator maths |
| "Security review is still open" | Security cleared two weeks ago; the note wasn't updated because there's no reason to rush it |
| "I want one more verbal before I commit it" | Verbal already happened; the rep is deciding which period benefits more |
| "Legal's slow on their end" | Legal returned redlines days ago |
If you only ever look at the forecast number, this reads as caution. If you pull the activity timeline against the CRM stage-change log, it reads as scheduling.
The fix is the plan, not the person
You cannot coach your way out of a structural incentive. You can only remove or blunt it. In order of what actually moves the needle, based on what I've seen work:
- Decouple next period's quota from this period's actuals, or at minimum cap the ratchet — no more than a 10% step regardless of over-attainment. This alone removes the single biggest reason to hold deals across a quarter boundary.
- Smooth the accelerator instead of stepping it. A continuous curve from 90–150% removes the cliff that makes timing valuable. Reps stop caring exactly when a deal lands because the maths barely changes either way.
- Kill the decelerator, or replace the cap with a SPIFF. If you're worried about windfall payouts, pay a bonus, don't punish the marginal deal. A capped or decelerating rate guarantees under-reporting near the ceiling.
- Extend true-up windows transparently, rather than leaving them as an informal, word-of-mouth escape hatch reps discover from their manager instead of from the plan document.
If you're not sure whether your current quota is even a fair target to begin with — which is often the real question hiding underneath a sandbagging conversation — that's worth checking before you touch the accelerator at all; the "Is My Quota Actually Winnable?" Self-Assessment is built for exactly that gut-check. And if a rep genuinely can't tell you what their own plan pays them at 110% versus 140%, don't blame them for optimising around a document they've never fully parsed — most haven't; that's what the Commission Statement Decoder exists to fix.
What I'd actually ask in the forecast call
Not "why isn't this in commit." Ask: "If you closed this today instead of next Tuesday, what changes for you financially?" If the honest answer is "nothing, really," you've found your sandbagging cause in one question, and it isn't the rep's character. It's a line item in the plan you wrote.
The reps who game accelerators aren't your weakest performers. They're usually your best ones — the only people paying close enough attention to the compensation structure to notice the incentive exists. Punish that and you've taught your top quartile to be quieter about it, not less rational.