Your Comp Plan Can't Tell Skill From Luck. Most Territories Can't Either.
Two reps land the same number — one because they mastered discovery, the other because their patch had three renewals lined up. The plan pays them identically.
Sarah closed £340,000 last quarter. So did Dave. Same product, same 18% commission rate, same accelerator kicking in over 100%, same seat booked at President's Club. Pull the pipeline apart and the two numbers were built by two completely different processes. Sarah ran forty cold discovery calls, disqualified eleven accounts inside the first five minutes because the budget wasn't real, and closed six deals where she'd changed the buyer's mind about what problem they actually had. Dave inherited a book with three accounts already committed to a renewal uplift and one existing customer mid-way through an expansion conversation his predecessor had started. He answered the phone, signed the paperwork, and hit target in April.
The comp plan cannot tell these two reps apart. It was never built to. It measures the output because the output is the only thing that's cheap to audit — a number lands in a CRM field, gets multiplied by a rate card, and a cheque goes out. Whether that number came from six months of grinding discovery-call practice or three lucky renewals sitting in a territory nobody had touched in two years is invisible to the plan. And that's the problem, because the plan is also the mechanism you use to decide who gets promoted, who gets the next big account, and who you quietly let walk when the market gets tight.
Outcomes are cheap to audit. Skill is not.
Finance likes comp plans because they're auditable in a way almost nothing else in a sales org is. Revenue landed, multiply by rate, pay it. Nobody needs to sit in on a call to verify a number in Salesforce. That's exactly why outcome-only pay persists even though every VP of Sales I've worked with will admit, off the record, that territory quality swamps individual skill in at least a third of their book.
The audit trail for skill is harder to build. It requires someone watching or listening to actual selling behaviour — how a rep runs discovery, how they handle the objection that isn't in the playbook, whether they can read a buying committee instead of pitching to whoever picked up the phone — and scoring it against a standard that holds still from rep to rep. Most orgs don't have that infrastructure, so they don't measure it, so it doesn't exist in the compensation conversation. Not because it isn't real. Because it's inconvenient.
What a hot patch actually buys a rep
Territory luck isn't one thing. It's usually some combination of:
- Inherited pipeline — deals already in motion when the rep took the patch, closing on momentum somebody else built.
- Renewal density — a book weighted toward accounts up for renewal in the measurement window, which close near-automatically if nobody actively churns them.
- Market tailwind — a vertical or region where demand spiked for reasons that have nothing to do with selling skill.
- Account concentration — one or two large logos that would have expanded regardless of who held the pen.
None of these show up as a line item on a commission statement. They show up as a number that looks exactly like the number a skilled rep produces from nothing. Run one of your own statements through a Commission Statement Decoder and you'll see the gap — most show gross payout and quota attainment and absolutely nothing about pipeline provenance.
The tell: what happens when the luck runs out
Here's where it costs you. Dave's patch gets rebalanced next year, or the renewal wave passes, or the market cools. His number drops 40%. Nobody in the leadership team can explain why, because the only data anyone tracked was the outcome, and the outcome doesn't carry a "why" field. Two things typically happen next, both bad. Either the org concludes Dave has "lost his edge" and manages him out, when actually he never had much of an edge to lose — he just never developed one, because the plan never asked him to. Or worse: the org concludes the whole territory model needs an overhaul, and burns a quarter re-carving patches instead of building the coaching infrastructure that would have caught this in month four of Dave's tenure.
Meanwhile Sarah, who built her number the hard way, gets treated identically to Dave in every conversation that matters — comp reviews, promotion panels, retention-risk assessments — right up until she's poached by a competitor who actually asked her in the interview how she runs discovery, and she realises for the first time that someone outside her own company noticed the difference.
What actually separates the two, and how you'd know
| Signal | Skill-driven number | Territory-driven number |
|---|---|---|
| Pipeline self-sourced vs. inherited | Majority self-sourced | Majority inherited or renewal |
| Discovery call quality, scored against a standard | Consistently high across accounts | Variable, often skipped on "easy" accounts |
| Deal cycle on genuinely new logos | Comparable to team average or faster | Long or nonexistent — no new-logo reps to compare |
| Performance if the patch were swapped tomorrow | Would likely persist within 1–2 quarters | Would likely collapse |
| Coachable pattern visible across multiple deals | Same technique, different accounts | Each win has a different, external explanation |
The only row on that table the comp plan actually measures is none of them. It measures the total at the bottom, after all five variables have already been baked in and made indistinguishable.
What to do that isn't "redesign the whole comp plan"
You don't need to blow up variable pay to fix this. You need a second, independent measurement running alongside it. Score the behaviour, not just the outcome, using a standard that's consistent across every rep and every account, and put that score next to the commission statement in every 1:1, every promotion conversation, and every territory reshuffle. A Sales Competency Self-Rating Scorecard is a blunt but honest starting point if you have nothing else running — it at least forces the conversation to separate "I hit number" from "I got better at the thing that produces numbers." Pair it with an honest look at the territories themselves; a Territory Design Fairness Scorecard will tell you, before the comp cycle even starts, which patches were never a fair comparison in the first place.
None of this replaces the comp plan. It gives you the thing the comp plan structurally cannot produce on its own: an answer to "was that skill, or was that the patch," asked while it still matters — not eighteen months later, in an exit interview, from a rep you already lost.