Compensation benchmarking
Compensation benchmarking is comparing what a company pays a role against what the market pays for comparable work, usually expressed as a percentile of survey data. A company targeting the 50th percentile intends to pay around the middle of its market; targeting the 75th means paying above three quarters of comparable employers.
Survey data answers a narrower question than the one most companies ask of it. The survey can say what a set of participating employers reported paying for a role they matched to a benchmark description. It cannot say whether your version of that role is the same job. Those two statements are close enough that the gap between them is easy to miss, and almost every bad benchmarking decision lives inside it.
The mechanism is percentile positioning. A survey reports pay at the 25th, 50th and 75th percentiles for a role, usually cut by company size, industry and geography, and a company picks where it intends to sit. The 50th is the default and means roughly market rate. Above it is a deliberate spend: paying at the 75th costs more per head and is generally justified by lower attrition, a shorter time to fill, or a genuine scarcity in the skill. Below it is also a strategy, and an honest one when the company is offering equity, seniority or a market position that reps want. What it cannot be is an accident, which is what it becomes when nobody looks for three years and inflation does the positioning instead.
For sales roles the title match is where it goes wrong. Account Executive spans a rep closing $15,000 transactional deals in a four-week cycle and a rep running $1.4 million enterprise pursuits over nine months, and both get reported under the same benchmark. Matching on scope produces a usable comparison: quota carried, average deal size, segment, sales cycle, whether the role is a hunter or covers an existing book. The other half of the problem is pay mix. A $130,000 on-target figure at 80/20 puts $26,000 at risk and one at 50/50 puts $65,000 at risk, and a survey that reports only the on-target total presents those as the same job at the same pay. Compare base to base and variable to variable, or compare nothing.
A team benchmarked against its own job title
A company benchmarks 26 account executives against survey data for the same title, in the same region and company-size band. Its reps carry a $1,400,000 quota; the survey's median respondent carries $850,000.
- Survey 25th percentile on-target earnings
- $105,000
- Survey 50th percentile
- $130,000
- Survey 75th percentile
- $162,000
- The company's current average
- $121,000
- Interpolated position between the quartiles
- about the 41st percentile
- Quota carried, company against survey median
- $1,400,000 against $850,000
- Cost of moving 26 reps to the 50th percentile
- $234,000 a year
The $9,000 a head is the easy part of this. The harder finding is the last row: these reps carry roughly 1.6 times the quota of the survey population, so the median of that population understates the market rate for the job these reps are actually doing.
Benchmarking the package and ignoring the quota behind it
Two companies pay an identical $130,000 on-target. One sets quota so that two thirds of the team clears it; the other sets it so that a quarter does. The first company's reps earn close to target most years and the second company's earn well below it, so the advertised figures match and the paychecks do not. A benchmark built on on-target earnings alone compares intentions rather than outcomes. Where the survey reports it, actual earnings and the share of the population reaching quota are worth more than the target figure, and where it does not, the target number needs reading alongside your own attainment distribution.
- What does the 50th percentile mean in compensation benchmarking?
- It is the midpoint of the reported market data: half of the comparable roles in the survey are paid more and half are paid less. A company targeting the 50th percentile is aiming to pay around the market rate for that role. Targeting the 75th means paying more than three quarters of the market, which is a deliberate spend usually justified by retention, hiring speed or scarce skills.
- How often should compensation benchmarking be done?
- Most companies run a full exercise annually, timed to land before the pay review and the next year's plan design. Survey data is collected at a point in time and ages, so a figure more than a year old understates a moving market. Roles in a hot skills market, or in a region where pay is shifting quickly, are often checked more frequently than that.
Nearby terms
Knowing the word is
the easy half
Commish pays compensation benchmarking the way your plan describes it, shows the arithmetic on every line, and traces each payment back to the deal that earned it.