How Do I Keep Reps From Gaming the Comp Plan?

The two anti-gaming architectures, compared

Strip away the vendor logos and every fix for Gaming a Comp plan falls into one of two families. The first is plan-math hardening: you keep paying commission, but you change what the math rewards. You add margin components, attach-rate multipliers, renewal gates, clawbacks, payout caps, accelerators that only unlock above a quality threshold. The check is still a commission calculation — it just has more inputs. The second family is scorecard and visibility: you stop treating pay as the only signal and build a weighted matrix of the full job, score every rep on every line, publish the standings, and wire the coaching cadence (and often the payout) to that composite.

They are not rivals. They attack different halves of the same problem. Plan-math hardening removes the *financial* payoff from a gamed behavior. Scorecard visibility removes the *social and career* payoff — a rep who juices one metric now watches their composite drop in front of the whole floor, and their manager's one-on-one opens with it.

Here is the honest split. Plan-math hardening is necessary but slow. Changing a comp plan mid-year is a contractual event; most orgs do it once or twice a year, and every change costs you trust with the field. Reps know this, which is why they treat a plan as a fixed puzzle to solve. Scorecard visibility is fast — you can re-weight a matrix in an afternoon and the incentive landscape shifts by Monday — but it is softer, because visibility alone does not move money unless leadership actually uses it in reviews and promotion decisions.

The strongest programs run both, sequenced. Build the matrix first so you have a shared, defensible definition of what "good" means. Then use the matrix to decide which plan components deserve real commission weight. Teams that skip the matrix and go straight to plan surgery usually end up with a baroque commission document nobody can explain, and complexity is itself a gaming surface — reps exploit the parts their manager does not understand.

A third pattern shows up constantly and deserves naming because it is not really a fix: surveillance. Dashboards, call recording, activity scoring, stage-hygiene alerts. These are useful inputs, but they only catch Gaming after it happens and only if someone acts on the flag. Surveillance without a scoring model produces a pile of alerts and a manager who stops reading them by week three.

There is also a fourth, quieter architecture: remove the incentive to game by removing the cliff. A lot of Gaming is not malice, it is a rep three deals short of quota in the last week of the quarter doing arithmetic. Smooth the payout curve, kill the binary accelerator cliff, pay on a rolling period, and a meaningful slice of Gaming evaporates because the math no longer punishes honesty. It is the cheapest fix available and the most frequently overlooked.

How to decide between them

The decision tree is deliberately unglamorous. Most teams jump to tooling before they have answered the first question — *which specific behavior am I trying to stop?* "Gaming" is not one behavior. Sandbagging a forecast, discounting to close, stuffing pipeline with junk, front-loading the easy SKU, parking deals in a stage, splitting a deal to double-count, and dumping a renewal into next period are seven different problems with seven different fixes. A weighted scorecard deters most of them because each one shows up as a weak level on a specific line. A payout cap deters only the ones that produce a spike.

Concrete numbers behind each option

Numbers make the trade-offs real. Below are ranges practitioners actually work with, not vendor promises.

Scorecard construction. Eight to nine KPI lines is the practical sweet spot. Fewer than six and you have not covered the job; more than twelve and managers stop filling it in honestly, which is worse than no scorecard because it manufactures false confidence. Common lines: bookings, gross margin percentage, discount discipline (deals closed within approved band), attach or multi-product rate, renewal and gross retention, forecast accuracy, pipeline hygiene or stage aging, and a qualitative or activity line. Weights typically run 15-25% on the two or three lines that carry the business, 5-10% on the supporting lines, and must sum to 100%.

Scoring scale. A 1-to-5 level per line is standard because it is coarse enough to be defensible and fine enough to differentiate. Define each level with a written anchor — level 3 means "meets the standard," level 5 means "top decile, and here is the number that proves it." Undefined levels are where scorecards die; two managers scoring the same rep three points apart destroys credibility faster than any gamed metric.

Composite math. Composite = sum of (weight × level). With nine lines and a 1-to-5 scale, the theoretical range is 1.00 to 5.00. In practice, a healthy team clusters between 2.8 and 4.2. A rep who is a 5 on bookings but a 1 on margin, attach, and forecast accuracy lands somewhere near 2.6 — visibly below a rep with no standout line but steady 4s. That single arithmetic fact is the whole anti-gaming mechanism, and it is worth walking the floor through the math once, out loud, with real examples.

Payout design. Accelerators that kick in at 100% of quota are the classic gaming magnet. Moving the first accelerator to 110% or 120% of quota, or making the accelerator apply only to margin-qualified revenue, removes the last-week scramble. Payout caps are blunt but effective — capping commission at, say, 250-300% of target variable pay closes the "one giant gamed deal" scenario, at the cost of occasionally capping a genuinely heroic quarter. Clawbacks tied to 90- or 180-day retention on new logos deter the discount-and-run play.

Forecast accuracy as a scored line. This is the highest-leverage single addition for most teams. Score it as absolute variance between committed forecast and closed result, per rep, per quarter. A rep who commits $400K and closes $250K is not "unlucky" — they are a level 1, and if forecast accuracy carries 15% weight, that alone costs them roughly 0.6 points on a 1-to-5 composite. Sandbaggers get caught by the same line from the other direction: consistent under-commitment is equally visible.

Discount discipline. Define approved discount bands by deal size and segment — for example, 0-10% self-approved, 10-20% requires manager, 20%+ requires a margin justification on record. Then score the percentage of a rep's closed deals that fell inside their band. Reps who habitually trade price for close show up immediately, and the scorecard gives the manager a factual opening instead of an accusation.

Administration cost. A spreadsheet scorecard costs nothing but roughly two to four hours a month of RevOps time to maintain and reconcile. Purpose-built incentive-compensation platforms commonly run from around $15 per user per month at the low end to enterprise contracts in the tens of thousands annually. Scorecard-and-coaching platforms sit in between. The admin cost is real either way — budget for it, because an unmaintained scorecard is a gaming surface of its own.

Time to effect. Plan changes take a full quarter or more to change behavior, because reps need to believe the new plan is permanent. Scorecard visibility changes behavior in two to four weeks if the standings are published weekly and managers actually reference them. That speed difference is the strongest argument for building the matrix before you touch the comp document.

Implementation details and sequencing

Sequence matters more than tooling. Start by naming the specific behaviors you want to stop, in plain language, with examples from your own floor. "Reps are holding deals to next quarter to protect a club trip" is a usable problem statement. "Reps are gaming the plan" is not.

Draft the lines and the anchors before you set weights. Anchors are the load-bearing wall. Write, for each line and each level, the observable standard. If two managers cannot independently arrive at the same level for the same rep, the anchor is too vague.

Set weights with finance in the room. This is the step teams skip and regret. RevOps can design a beautiful matrix, but if finance does not recognize the composite as a legitimate predictor of healthy revenue, the scorecard never earns authority in planning conversations. Bring them in early and let them argue about weights — that argument is the alignment.

Back-test before you launch. Score last quarter retroactively for every rep. Two things happen. First, you find out whether the anchors work. Second, you find out whether the composite ranking matches what managers already believe about their teams. Where it does not, the matrix is usually wrong, not the manager. Fix the anchors and re-run.

Publish, then explain the arithmetic on the floor. Walk through a worked example with real (anonymized) numbers. Show how a big-bookings, low-margin rep lands below a steady performer. Reps are numerate; once they see the composite respond to the whole job, the incentive to juice one line drops sharply.

Wire it to something that matters within 30 days. Weekly standings on a screen. The lowest-scoring line opens every one-on-one. Promotion and territory decisions reference the composite. If nothing happens for a quarter, reps correctly conclude the scorecard is theater and go back to the old behavior.

Then, and only then, change the comp plan. Use the matrix to decide which lines deserve real commission weight. The scorecard tells you which behaviors are chronically weak across the team; those are the ones worth paying for. This is the sequencing most teams invert, and inverting it is why so many comp-plan rewrites fail.

Re-weight quarterly, and treat re-weighting as routine rather than punitive. When a new pattern appears — everyone parking deals in a single stage, or a sudden run of tiny low-margin deals — adjust the relevant weight and announce it. The announcement is the deterrent. Reps who understand the weights are a moving target stop investing in finding the next exploit.

One more sequencing note: run this alongside pipeline hygiene work, not after it. Scorecards depend on clean CRM data, and Gaming frequently shows up first as dirty data — stages that never age, close dates that slide, amounts that get edited late in the quarter. If your stage definitions are loose, fix them in parallel, because a scorecard built on unreliable inputs produces unreliable scores and loses the floor's trust immediately.

Related questions

How many KPIs should a rep scorecard include?

Eight to nine lines is the working range. Fewer than six leaves parts of the job unscored — and unscored behavior is where Gaming hides. More than twelve and managers fill it in carelessly, which is worse than no scorecard. Cover bookings, margin, discount discipline, attach, retention, forecast accuracy, pipeline hygiene, and activity.

Should reps see the weights?

Yes. Publish the weights and the level anchors. Secrecy invites conspiracy theories and turns every score into an argument. Transparency gives reps a roadmap: they can see exactly which line is dragging their composite and work on it. That is the behavior you want, and it is only available if the math is visible.

What if a rep is excellent at one thing and weak everywhere else?

The weighted composite handles it automatically. A rep who is a 5 on bookings but a 1 on margin, attach, and forecast accuracy lands around 2.6 on a 1-to-5 scale — below a steady 4s performer. The payout follows the composite, so the specialist gets a clear, non-negotiable signal to broaden.

Can reps game a weighted scorecard too?

Eventually they will probe it, yes. But the matrix is far harder to game than a single metric because improving one line costs nothing on the others only if the underlying work is real. And because you control the weights, a new pattern gets neutralized in an afternoon rather than at the next annual plan rewrite.

How often should weights change?

Re-weight when a new gaming pattern appears or when business priorities genuinely shift — realistically once or twice a year for structural changes, plus ad hoc adjustments when you spot something. Announce every change. Frequent silent re-weighting destroys trust as fast as a stale matrix destroys relevance.

FAQ

What counts as Gaming a Comp plan, exactly? Any behavior that improves a measured number without improving the underlying business. Sandbagging forecasts, discounting to close, stuffing pipeline with deals that will never close, front-loading the easiest product, splitting deals to double-count, and pushing renewals into the next period are the common ones. The test is simple: would this behavior still be worth doing if nobody were measuring it?

We already pay on multiple components. Why are Reps still Gaming? Multi-component plans help, but if one component dominates the payout — usually bookings — reps optimize that one and treat the rest as noise. Check the actual payout distribution: if the top quartile of earners got there almost entirely on one line, your plan is effectively single-metric no matter how many components it lists.

How do we handle a rep who games but still hits quota? Score them on the full matrix and let the composite speak. A rep with big bookings and poor margin, attach, and forecast accuracy is not a top performer; they are a specialist with a visible weakness. Put the lowest line at the top of their next one-on-one and give them a quarter to move it. If the behavior persists, it becomes a performance conversation, not a comp conversation.

Does this work for SDR teams and renewals teams, not just closers? Yes, and the shape barely changes. For SDRs, swap bookings for qualified meetings held, meeting-to-opportunity conversion, and data quality. For renewals, weight gross retention, expansion, and forecast accuracy. The principle is identical: never let one number carry the whole check.

What is the single fastest change we can make this quarter? Add forecast accuracy as a scored line with meaningful weight, and publish it weekly. Sandbagging and pipeline theater are the most common forms of Gaming and the most expensive, because they corrupt planning for the entire company. It costs nothing to add and it changes behavior in weeks.

How does RevOps keep the scorecard from becoming shelfware? Tie it to something with consequences within thirty days — coaching agendas, promotion decisions, territory assignments, or a portion of variable pay. A scorecard that produces no visible consequence for a full quarter is dead, and reps will correctly ignore it.

Sources

• Salesforce — sales performance and quota management
• HubSpot — sales compensation and quota planning resources
• Harvard Business Review — research on incentive design and unintended consequences
• MIT Sloan Management Review — compensation and motivation research
• Gartner — sales compensation and revenue operations research
• WorldatWork — compensation program design and governance
• SHRM — incentive pay and total rewards guidance
• Corporate Finance Institute — sales compensation structures explained

Related on PULSE

• [How do we fix comp when reps are gaming deals, inflating pipelines, and sandbagging?](/knowledge/q278)
• [How should I structure SDR commission to discourage gaming MQL counts?](/knowledge/q02)
• [What quota credit policies prevent gaming while rewarding split deals and expansions?](/knowledge/q731)
• [How do you keep remote reps accountable without micromanaging?](/knowledge/q14027)
• [How do you coach reps to keep their CRM clean and current?](/knowledge/q13942)