Design Compensation Structure

---
name: design-compensation-structure
description: Use when building or redesigning a salary band structure, total rewards framework, or compensation philosophy for an organization
source: WorldatWork total rewards framework; Radford compensation survey methodology; SHRM compensation management guidelines
tags: [hr, compensation, total-rewards, salary, talent-management]
verified: true
---

# Design Compensation Structure

Build a defensible, market-aligned compensation structure that attracts, retains, and motivates talent.

## Why This Is Best Practice

**Adopted by:** Mercer, Aon Hewitt, Radford-aligned organizations, and enterprises using structured compensation bands
**Impact:** WorldatWork surveys show organizations with defined salary bands have 18% lower compensation-related turnover and 22% higher offer acceptance rates. Pay equity audits in banded structures identify and close gaps 3x faster than unstructured pay.
**Why best:** Salary bands create internal equity (similar roles paid similarly), external competitiveness (anchored to market data), and career progression clarity (movement through bands signals growth).

Sources: WorldatWork "Total Rewards Model" (2021); Radford Global Technology Survey methodology; SHRM Compensation Management Guide (2023)

## Steps

1. **Define compensation philosophy** — establish the organization's market position: lead (75th percentile), meet (50th percentile), or lag (25th percentile) market. Document the rationale and who it applies to (all roles, technical roles only, executives).

2. **Choose a market data source** — select 1–3 compensation surveys appropriate to industry and geography (Radford, Mercer, Culpepper, Levels.fyi for tech, Glassdoor/LinkedIn for directional). Match jobs to survey benchmarks, not titles.

3. **Define job architecture** — create a level framework (e.g., IC1–IC6, M1–M4) with clear scope, impact, and independence criteria for each level. This is the skeleton the bands hang on.

4. **Build salary bands** — for each level, define minimum, midpoint, and maximum. Standard band width is 50–80% (max ÷ min − 1). Midpoint should equal your target market percentile.

5. **Ensure band overlap** — adjacent levels should overlap by 25–40%. Overlap allows a top-performing L3 to earn more than an entry L4, preventing forced promotions to get pay increases.

6. **Add variable compensation** — define bonus targets as % of base salary by level (e.g., IC: 0–15%, M3: 20%, VP: 30%). Document performance linkage, payout timing, and discretionary vs. formulaic structure.

7. **Design equity component** — define equity eligibility by level, grant size ranges (as % of comp or dollar value), vesting schedule (4-year/1-year cliff is standard), and refresh grant policy.

8. **Run a pay equity audit** — plot current employee pay against new bands by role, level, gender, and ethnicity. Identify employees below band minimum (must fix immediately) and compression issues.

9. **Build merit increase framework** — define annual merit budget allocation matrix: % increase based on performance rating × position in band (compa-ratio). Employees below midpoint should receive larger increases to accelerate market alignment.

10. **Document and communicate** — write a compensation philosophy document. Share band ranges with employees (transparency reduces rumors and increases trust). Train managers on how bands work and how to have pay conversations.

## Rules

- Never set bands without external market data — internal equity alone creates market disconnects.
- Fix employees below band minimum before the next performance cycle; this is a legal risk in jurisdictions with pay equity laws.
- Review bands annually — markets move 3–8% per year; stale bands compress tenure-based pay.
- Keep band midpoints as the anchor, not the ceiling — employees should be able to reach the maximum through sustained high performance.
- Variable pay should be meaningful — a 3% bonus at 100% target is not a motivator.

## Common Mistakes

- **Using only one data source** — single-survey anchoring misses market variation; blend 2–3 sources.
- **Title matching instead of job matching** — a "Senior Engineer" at a startup is not the same as at a FAANG; match on scope and responsibilities, not title.
- **Ignoring geographic differentials** — San Francisco and Austin have 20–35% cost-of-labor differences; location-based banding is required for distributed teams.
- **No refresh grant policy** — early equity vests out; without refreshes, retention incentive disappears at year four.

## When NOT to Use

- Fewer than 20 employees (informal individualized negotiation is more appropriate)
- Pure commission sales roles where total comp structure is driven by output (use a different commission plan design process)
- Highly specialized roles with fewer than 3 market survey matches (use individual market pricing instead of bands)