sales_compensation_design v1.0.0

name: "sales_compensation_design" description: "Design sales pay plans that pay for the right sales: set on-target earnings and pay mix by role, pick three or fewer weighted measures, set quotas against a healthy attainment distribution, build the threshold to excellence payout curve with accelerators and a decelerator instead of a hard cap, write crediting, draw, ramp, and clawback rules that define when pay is earned, model team cost before launch, and run the yearly review. Use when creating or fixing a commission or quota plan, setting OTE or pay mix, adding accelerators, caps, or clawbacks, costing a plan change, or diagnosing sandbagging, windfalls, or a plan everyone games."

Sales Compensation Design

Purpose

Build a sales pay plan a rep can explain back to you and finance can cost before it launches. Most plan failures are mechanical, not motivational: too many measures, a quota set from the budget instead of the territory, a curve that pays the same slope everywhere, a cap that tells top reps to stop, or a clawback clause that tries to recover pay the plan already called earned. This skill designs the plan in order: role, earnings, measures, quota, curve, crediting, legal earning rules, cost model, then the yearly review. For pricing base pay against the market, use a compensation benchmarking skill. This one designs the variable pay on top of it.

Workflow

1. Start from the role, not the spreadsheet. Write what the role actually controls: new logos, expansion, renewals, or a blend, plus sales cycle length, deal size spread, and how much of the result is the rep versus the territory, the brand, or an overlay team. Pay mix follows influence and variability. A role the rep barely influences gets a salary heavy mix or a team bonus, not a personal quota plan. A true sales role should still carry real money at risk. One published practice assessment puts the floor at 85/15 base to target incentive for a sales role, and anything softer than that belongs in a bonus plan, not a sales plan.

1. Set on-target earnings, then split the mix. On-target earnings (OTE) is base plus target incentive at exactly 100 percent of quota. Set OTE from the market and the role value first, then choose the mix. Working patterns to argue from, not rules: new account hunters often sit near 50/50, blended sellers near 60/40, current account farmers near 70/30, first line managers lean more to base than their reps. A richer incentive share gives more leverage and more income risk, and it costs the company more when the team beats target, because accelerators apply to a bigger target incentive. State the mix per role in writing. Two roles with the same OTE and different mixes are different jobs on paper and should be.

1. Pick three or fewer measures, and weight them on purpose. The WorldatWork and OpenSymmetry sales compensation survey reported by SHRM found 71 percent of responding organizations use three or fewer performance measures. More than three and reps stop steering by the plan and start steering by whichever measure is easiest. Weight measures by the time and result you actually want: if total revenue should take about three days a week and new account revenue two, a 60/40 weighting states that plainly. Keep one primary volume measure (revenue, bookings, or collected cash). Add a second measure only for a real strategic trade off, such as margin, new logos, or retention. Anything that is simply expected of the job, such as CRM hygiene or activity counts, stays a job expectation or a gate, not a paid measure.

1. Set quota from territory potential, then test the distribution. Build quotas bottom up from territory potential, pipeline, and headcount capacity, then reconcile to the company target. Never set quota by dividing the budget by headcount and adding a growth wish. Then test the plan against a healthy attainment distribution. Aon Radford guidance: when the company hits its target, a healthy plan has roughly 50 to 60 percent of reps at or above quota, about 10 percent at or above the excellence level, and about 5 percent below threshold. If your forecast puts 90 percent of reps over quota, quotas are soft and accelerators will blow the cost model. If it puts 80 percent under, quotas are fiction and the curve will not motivate anyone. Define three points by name before building the curve: threshold (minimum acceptable performance), target (100 percent of quota), and excellence (the level a top performer, roughly the top tenth, can reach).

1. Build the payout curve as slopes, not a single rate. The base rate is target incentive divided by quota. Worked anchor: a 40,000 target incentive on an 800,000 goal is a 5 percent base rate. A single flat rate pays too much for weak performance and too little for excellence, so shape the curve in segments, using the bundled tool so every number in the plan document comes from the same math:

bash
   python3 scripts/payout.py table --ote 240000 --base-pct 60 --quota 2400000
   

The default curve in the tool pays nothing below a 50 percent threshold, 25 percent of target incentive at threshold, 100 percent at target, 200 percent at 150 percent attainment (excellence), and 250 percent at 200 percent. Its slopes tell the design story: 1.5 times payout slope from threshold to target, 2.0 times from target to excellence, then 1.0 times past excellence. That last segment is a decelerator, a regressive rate that keeps paying without paying runaway rates on a possible windfall. On a 240,000 OTE at a 60/40 mix (144,000 base, 96,000 target incentive, 2,400,000 quota, 4 percent base rate) the tool gives: 60,000 incentive at 75 percent attainment, 96,000 at 100 percent, 134,400 at 120 percent, 192,000 at 150 percent, and 240,000 at 200 percent. Excellence should pay about 2 to 3 times target incentive. If your curve pays 1.2 times at excellence, top reps will leave. If it pays 5 times at a level many reps reach, finance will kill the plan mid year.

1. Choose threshold and cap mechanics with your eyes open. A hard threshold pays nothing below it. It is clean and brutal, and on an annual quota with a low base it can leave a rep with no incentive for months. A soft threshold pays a discounted rate below a point instead, for example a reduced rate under 70 percent of quota, which keeps some pull on weaker performers. Cumulative year to date measurement softens both, because a bad quarter can be recovered. Prefer a decelerator over a hard cap. A cap tells your best rep to stop selling, defer deals, or leave, and everyone can see it. If a cap is unavoidable, for low margin or windfall prone deals, set it high, pair it with a written windfall clause for single deals that dwarf a territory, and say in the plan what happens to deals past the cap. Never hide a cap inside a vague management discretion clause.

1. Write crediting rules before the first deal closes. Cover, in writing: who gets credit when two reps split a deal and what the splits sum to, overlay credit and whether overlay credit is additive cost or carved from the primary rep, new business versus renewal versus expansion rates, multi year deals (credit on total contract value, first year value, or cash collected, and at what discount for prepay), team versus individual credit for managers, and what happens on territory changes mid year. Double crediting is a cost decision, not a favour. Price it in the cost model in step 9 or do not offer it.

1. Define when pay is earned, then write draws, ramp, and clawbacks to match. This is where plans become legal documents, so be exact and get local counsel to review before use. The plan must state the earning trigger in plain words: booked, signed, delivered, invoiced, or collected, and any window that must pass first. A clawback that tries to take back pay the plan already defined as earned is the classic failure. California shows the shape of the law in many places: Labor Code section 2751 requires written commission agreements, section 221 bars an employer from collecting back wages already paid, and case law upholds chargebacks only where the payment was an advance and the written plan defined the commission as not earned until its conditions, such as a cancellation window, had passed. So design it that way everywhere: pay early amounts as labelled advances, define earning as after the return, cancellation, or collection condition clears, set a short and stated chargeback window, charge back only the rep who earned the credit, and never let a chargeback cut base pay or take pay below minimum wage. Keep listed company executive clawbacks separate in your head: SEC Rule 10D-1, adopted in 2022 under Dodd-Frank section 954, forces recovery of erroneously awarded incentive pay from executive officers after an accounting restatement, with a three year lookback and no fault needed. That rule governs executives of listed companies. It is not a licence to claw back field commissions, and field plan language should never cite it as one. Draws follow the same logic: state whether a draw is recoverable against future incentive or a non recoverable guarantee, the amount, and the exact month it ends. Ramp follows it too: state the ramp length, the reduced quota or guaranteed payout during ramp, and the date full quota starts. A guarantee with no end date becomes the plan.

1. Cost the plan on a distribution, never on everyone at 100 percent. A plan where every rep lands exactly at target never happens, and costing it that way understates the bill because accelerator payments to top reps exceed the savings from weak reps. One practice benchmark budgets sales compensation at about 108 to 110 percent of the sum of target total compensation for that reason. Model your real expected spread with the tool:

bash
   python3 scripts/payout.py cost reps.csv --ote 240000 --base-pct 60 --quota 2400000
   

Input columns: rep,attainment_pct. Worked run, ten reps at 45, 60, 75, 85, 95, 100, 100, 115, 130, and 150 percent attainment on the plan in step 5: total bookings 22,920,000, total incentive 924,000 (4.031 percent of bookings), total compensation 2,364,000, which is 98.5 percent of the sum of OTEs at 95.5 percent team attainment, with 50 percent of reps at or above quota and 10 percent at excellence. Run at least three distributions before launch: team at about 90, 100, and 110 percent of target. Report the incentive cost of sales for each. If the 110 percent case breaks the margin, fix the curve now, not in a mid year plan change.

1. Launch it like a contract, then shadow run it. Issue a written plan document per role using the template in references/templates.md, signed or acknowledged before the period starts. Include eligibility, measures, weights, quota, the full payout table, crediting, earning and chargeback rules, draws and ramp, payout timing, the dispute window and who decides, and the narrow conditions under which the company may amend the plan. Run the first payout cycle as a shadow calculation beside the old method or a hand check on the biggest and smallest payouts. Pay on a stated cadence, monthly or quarterly for most quota plans, close enough to the result that the rep connects the pay to the work. Late or unexplained statements destroy trust faster than a stingy curve.

1. Run the yearly review on evidence, not complaints. Work the checklist in references/worksheets.md. Read the attainment distribution against the healthy shape in step 4, the payout dispersion (compare the 90th percentile payout to the median against the leverage you designed), incentive cost of sales against the model in step 9, and behaviour signals: discounting to pull deals in, deals slipping past period end, sandbagging after quota, churn on accounts paid at new business rates, and disputes by cause. Separate three kinds of problem and fix only the one you have: quota problems (distribution is skewed for everyone), curve problems (pay moves too fast or too slow between points), and measure problems (reps hit the plan while the business misses its goal). Change plans at the year boundary. Mid year changes, except to fix a plain error or a legal problem, teach reps that the plan is negotiable and next year they will wait you out.

Output Contract

A plan design deliverable contains:

• A role sheet per role: what the role controls, OTE, base and target incentive, pay mix with the reason for it, and quota basis
• The measures table: three or fewer measures, each with definition, weight, data source, and measurement period
• Quota build: territory or capacity basis per rep or segment, company target reconciliation, and the forecast attainment distribution against the healthy shape (50 to 60 percent at or above quota, about 10 percent at excellence, about 5 percent below threshold)
• The payout curve: threshold, target, and excellence points, payout percent of target incentive at each curve point, segment slopes, cap or decelerator mechanics, and a payout table generated by scripts/payout.py at 0 to 250 percent attainment
• Crediting rules: splits, overlays, new versus renewal versus expansion, multi year treatment, manager credit, and mid year territory change handling
• Earning and recovery rules: the earning trigger, advance versus earned labels, chargeback triggers and window, draw terms, ramp terms, and the departures and leave cases from references/edge-cases.md
• A cost model: at least three team distributions (about 90, 100, and 110 percent of target) with total incentive, total compensation, and incentive cost of bookings for each, from scripts/payout.py cost
• A signed plan document per role from the template in references/templates.md, plus the first cycle shadow check result
• A yearly review record: distribution, dispersion, cost, behaviour signals, disputes by cause, and each change made with its reason

Operating Rules

• Never present a payout number that the bundled tool or a shown hand calculation did not produce. Curve points, slopes, and cost totals must reconcile with each other in the same document.
• Never call a payment earned and recoverable in the same plan. Pick the earning trigger, label early payments as advances, and write the window. Get local employment counsel to review earning, chargeback, draw, and final pay language before the plan is used. This skill is design guidance, not legal advice.
• Never set quota from the budget alone, and never promise a distribution shape as a fact about a specific team. The 50 to 60 / 10 / 5 shape is a diagnostic benchmark from published practice guidance. A real team's shape is evidence about its quotas, not a target to force.
• Do not add a measure to fix a behaviour the manager should manage. Paid measures are for results the role controls. Gates and job expectations handle the rest.
• Do not cap quietly, amend mid year casually, or pay on a cadence the document does not state. Surprises in pay read as dishonesty even when the math is right.
• Keep executive clawback law (SEC Rule 10D-1) and field commission chargebacks in separate sections of any document. Different people, different triggers, different law.
• If two measures can be gamed against each other, such as revenue paid at booking and retention paid a year later to a different rep, say who bears the trade off in writing before launch.