Our leveling system was sophisticated. 12 levels. Detailed rubrics for each. Technical skill, scope, impact, communication, leadership—all scored on 5-point scales. Scientifically designed.
The result: endless arguments about imaginary distinctions.
The 4.0 vs 4.5 Debate
Promo committee meeting. Two hours. The question: is Sarah a 4.0 or 4.5?
Manager A: "She led a cross-functional project. That's 4.5 scope."
Manager B: "But the project was within her team's domain. I'd say 4.0."
Manager A: "She influenced the design of Team X's integration."
Manager B: "Influenced, but didn't own. That's 4.25 at best."
We didn't have 4.25. We were debating decimal places on a system that didn't have them. The distinction between 4.0 and 4.5 was arbitrary. We pretended it was precise.
The Precision Illusion
More levels feel more accurate. But accuracy requires:
- Clear, objective criteria (we didn't have this)
- Consistent application across raters (we didn't have this)
- Meaningful differences between levels (we didn't have this)
Study after study shows manager ratings have low inter-rater reliability. Two managers evaluating the same engineer often disagree by 1-2 levels. Our 12-level system implied precision we couldn't deliver.
The Consequences
Political leveling: Employees spent energy arguing for level bumps instead of doing impactful work.
Inflation pressure: Managers wanted to reward team members. Only tool available: level bumps. Levels inflated over time.
Comparison anxiety: "She's a 6 and I'm a 5.5, but I'm better!" The granularity invited granular comparisons.
Promo bottlenecks: Each half-level required justification. Committees spent weeks on calibration.
The Collapse: 4 Bands
We replaced 12 levels with 4 bands:
- Developing: Learning the craft, needs guidance
- Contributing: Reliably productive, owns work independently
- Leading: Multiplies others, shapes direction
- Executive: Organizational scope, strategic decisions
That's it. No half-levels. No 5-point rubrics. No calibration spreadsheets.
How It Works
Band placement: One question—which band best describes your impact? Debates are rare because the bands are obviously different.
Compensation: Wide ranges within bands. You can get significant raises without changing bands. Growth without promotion theater.
Progression: Moving between bands is rare and obvious. Contributing → Leading is a real change in how someone works, not a checkbox exercise.
The Objections We Heard
"People need clear goals!" Goals should be about impact, not level. "Help two junior engineers become independent" is a goal. "Reach L5.5" is not.
"How do I know I'm growing?" Feedback from your manager. Increasing scope. More complex problems. Real signals, not arbitrary levels.
"How do candidates compare offers?" We share comp bands publicly. External recruiters can map to their levels if needed. Not our problem to solve.
Two Years Later
- Time spent on leveling discussions: Down 80%
- Compensation satisfaction: Up (wider ranges = more room for raises)
- Level-related complaints to HR: Near zero
- Regrets about simplifying: None
12 levels looks precise. 4 bands is honest. Honest beats precise when precision is fake.
Written by XQA Team
Our team of experts delivers insights on technology, business, and design. We are dedicated to helping you build better products and scale your business.