Variable pay and bonus plan design — what behavior are you actually buying?
Bonus plans are behavioral instruments. The choice between individual, team, and company performance pools tells employees what you actually value — usually…
On this page▾
- Variable pay is a signaling instrument first, motivation instrument second. The structure communicates what the company actually values regardless of stated values.
- Three pool architectures dominate: individual performance (sales-style), team performance (function-level), and company performance (profit-share). Most plans blend two of the three.
- Goodhart's Law is the dominant risk: any KPI tied to bonus becomes a target rather than a measure. Build in subjective performance modifiers to keep the metric honest.
- Bonus communication failure: announcing the bonus only at payment time. The signaling value of a bonus comes from the conversation that ties behavior to reward — not the deposit notification.
Most bonus plans are designed by reading what competitors do and adapting. The result: a behavioral instrument whose effects nobody traces. The question to design backward from is 'what behavior am I willing to pay extra for?'
What bonuses actually do
- Signal what the company actually values (regardless of what the values page says)
- Differentiate top performers from average (or hide that differentiation)
- Create a counterweight to base salary cost — if 20% of comp is variable, you can ramp it down in a bad year without layoffs
- Provide a discretion mechanism for managers to recognize outsized contribution
The three pool architectures
| Pool | Best for | Risk |
|---|---|---|
| Individual performance | Sales, BD, account management — anywhere individual output is measurable | Tournament behavior; weak collaboration |
| Team / function performance | Product, engineering, ops — where output is collective | Free-rider problem if team is large; political tension around team scope |
| Company performance (profit-share or company-wide bonus pool) | Late-stage / mature companies; back-office functions | Weak link between individual behavior and reward; bonus becomes deferred salary |
Most modern plans blend two: company performance gates the bonus pool size (no profit, no bonuses); individual or team performance determines distribution. This protects the business in down years while keeping individual signal in good years.
Goodhart's Law in practice
Sales rep paid on closed deals → reps stuff Q4 with low-quality contracts that churn in Q2. Support team paid on tickets-closed → reps close tickets prematurely. Engineering team paid on velocity points → points inflate by 40% in two quarters. Goodhart's Law: 'When a measure becomes a target, it ceases to be a good measure.'
The mitigation: pair every quantitative metric with a quality modifier (NPS for sales, CSAT for support, retro signal for engineering). The modifier is subjective and applied by a calibrated manager — which slows down gaming but doesn't eliminate it.
Designing the payout curve
- 50% goal hit → 50% payout
- 100% goal hit → 100% payout
- 150% goal hit → 150% payout
- Simple; easy to communicate
- Doesn't differentiate stretch performance
- <70% goal hit → 0% payout
- 100% goal hit → 100% payout
- 150% goal hit → 200% payout (accelerator)
- Higher motivation at the top
- Less protective at the bottom; harder to budget for
Frequently asked questions
What's a typical bonus target as % of base?
10–15% for IC/manager roles, 15–25% for director/senior leader, 30–50% for VP+. Sales roles are entirely different — usually 50/50 base/variable or 60/40, with no upper cap on commission.
Should bonuses be paid quarterly, semi-annually, or annually?
Sales: monthly or quarterly to tighten the behavior loop. Non-sales: annually with optional mid-year true-up. Quarterly non-sales bonuses tend to feel like noise once the amount is divided by 4.
What about discretionary bonuses?
Useful as a small (≤10% of plan) discretionary pool that managers can use to recognize specific moments. Beyond that, discretionary bonuses become the place where pay-equity audits find the largest unexplained gaps — discretion is where bias lives.
- Drive (Daniel H. Pink) — Riverhead
- Goodhart's Law (Goodhart, 1975) — Bank of England
- WorldatWork — Variable Pay Survey — WorldatWork
Read next
All playbooksHow to set pay targets, bands, transparency, and review cadence — before you have to negotiate a single offer.
Why sales comp is its own discipline — quota setting, OTE design, accelerators, clawbacks, and the behavioral traps that turn comp plans into anti-incentives.
The system around the review matters more than the review itself. A modern approach to goals, feedback, calibration, and the conversation.