How to Build Salary Bands That Are Actually Defensible
Quick Answer: Building a defensible salary band requires three inputs used in a specific sequence: internal job evaluation to establish the role's relative worth, external market benchmarking to anchor the band to what the market actually pays, and a documented pay positioning decision (your deliberate choice to lead, match, or lag the market). Without all three, your band is either internally inconsistent, market-disconnected, or impossible to defend when challenged.
Most salary bands are built backwards. An HR team looks at what the organization has been paying for a role, adds a few percentage points of buffer on either side, and calls that the band. Or they pull a single salary survey number, multiply by 0.8 and 1.2 to get a minimum and maximum, and publish the result without ever examining whether those numbers reflect the actual internal value of the role or the organization's competitive position in the market.
The problem with this approach is not that it is careless. It is that it produces bands that cannot be explained. When an employee asks why their salary is at the bottom of their band, or when a regulator asks how the salary range posted in a job advertisement was determined, "we looked at survey data and added a buffer" is not an answer that survives scrutiny.
A defensible salary band is one where every number can be traced back to a documented process with a clear methodology, current market data, and a stated positioning rationale. This guide walks through exactly how to build one.
CompBldr's Market Benchmarking module connects JESAP evaluation scores, blended survey data, and your pay positioning strategy to produce salary bands automatically. Try the free salary band calculator or book a 15-minute demo.
The Three Inputs Every Salary Band Needs
Input 1: Internal Job Evaluation Score
Before you touch any market data, you need to know the relative internal value of the role you are banding. This is what a job evaluation methodology produces. A point factor system like JESAP scores each role across compensable factors: knowledge required, problem-solving complexity, accountability and organizational impact, and working conditions. The total score tells you where this role sits in your internal hierarchy relative to every other role in the organization.
Why does this matter for salary bands? Because without internal evaluation, your bands are anchored only to external market data. That means two roles with very different internal scopes might end up in the same band because they happen to carry similar market rates. Or a role might end up in a band that does not reflect its internal complexity because the survey comparison was done by title matching rather than scope matching.
Internal evaluation provides the structural address of the role. Once you know a role scores 312 points on JESAP and maps to Grade 5, you know it should be banded alongside every other Grade 5 role regardless of job family. That is internal consistency.
Input 2: External Market Benchmarking
The second input is market data that tells you what organizations comparable to yours are paying for roles with equivalent scope, level, and function. This data comes from compensation surveys published by providers like Radford (Aon), Mercer, and Willis Towers Watson.
The market anchor for most bands is the P50, the 50th percentile of market pay for the role. This is the midpoint of your salary band. From there, you set the minimum and maximum by applying a range spread.
Critical point: the market matching must use your job architecture context, not just your job title. Matching a "Senior Engineer" to the survey's "Senior Engineer" by title is unreliable because titles are not standardized. Matching based on job family, grade, scope, level of autonomy, and organizational impact produces a far more accurate market anchor.
Using multiple survey sources and blending them with weights by job family produces the most reliable midpoints. A typical blend might weight Radford at 50%, Mercer at 30%, and WTW at 20% for technology roles, and shift those weights for finance or operations roles where different surveys have better coverage.
Input 3: Your Pay Positioning Decision
The third input is a deliberate organizational decision about where you want to position your pay relative to the market. This is your pay philosophy expressed as a concrete choice.
Common pay positioning strategies:
- Market-match (P50): Set your midpoint at the 50th percentile. You pay at the median for the market. This is the most common strategy for organizations that want to be competitive without overpaying.
- Market-lead (P75): Set your midpoint at the 75th percentile. You deliberately pay above median to attract and retain talent in competitive markets, particularly for high-demand roles.
- Market-lag (P25 or P40): Set your midpoint below median. Sometimes appropriate for organizations that compete on other dimensions (mission, stability, total rewards) or for roles where supply significantly exceeds demand.
- Differentiated positioning: Different percentile targets for different job families. Lead at P75 for engineering where talent is scarce, match at P50 for finance, lag at P40 for administrative roles. This is the most sophisticated approach and requires the most documentation.
The positioning decision must be documented. "We target P50 for all roles" is a policy. "We targeted P50 for this role because that reflects our compensation philosophy for this job family" is a defensible explanation. Without documentation, every future question about why the band was set where it was starts from zero.
Setting Band Width: How Wide Should a Salary Band Be?
Band width is the percentage spread from minimum to maximum. It is typically expressed as a percentage of the midpoint. A band with a 50% spread has a minimum at 80% of midpoint and a maximum at 120% of midpoint. A band with an 80% spread has a minimum at 75% of midpoint and a maximum at 125%.
Common conventions by grade level:
- Entry and developing roles (Grades 1 to 3): Narrower bands, typically 40 to 50% spread. These roles have less variation in scope and the career progression within the band is shorter before a grade change is appropriate.
- Mid-level roles (Grades 4 to 5): Moderate bands, typically 50 to 70% spread. Enough range to accommodate skill development and performance differentiation within the same grade.
- Senior and leadership roles (Grades 6 to 7): Wider bands, typically 60 to 80% spread. Senior roles have more variation in scope and the tenure within a grade before moving to the next level is longer.
Wider bands are not automatically better. A very wide band means that an employee can spend many years in the same role without approaching the band ceiling, which can mask the need for a grade change or a market adjustment review. A very narrow band means that high performers hit the ceiling quickly and you either have to promote them or find alternative compensation mechanisms.
Band Overlap: Should Adjacent Bands Overlap?
Most well-designed band structures allow for 20 to 30% overlap between adjacent grades. This means the maximum of Grade 4 extends slightly into the minimum of Grade 5. Overlap serves an important purpose: it allows an employee at the top of one grade to be paid comparably to someone at the entry point of the next grade, reflecting the reality that a highly experienced individual contributor may contribute as much as a newly promoted manager.
Zero overlap creates a visible pay cliff where every promotion results in a jump to the new band minimum, which may be significantly higher than current pay or may require a salary cut from the top of the old band. Excessive overlap (more than 40 to 50%) blurs the distinction between grades and makes grade placement less meaningful for compensation purposes.
Building the Band: The Calculation
Once you have your three inputs, the calculation is straightforward.
Step 1: Determine your market midpoint. For a Senior Engineer role at Grade 5, your blended survey data shows a P50 of $145,000.
Step 2: Apply your positioning decision. Your organization targets P50 for engineering roles. Midpoint = $145,000.
Step 3: Apply your range spread. For Grade 5, you use a 70% spread. Minimum = midpoint / (1 + spread percentage / 2) = $145,000 / 1.35 = approximately $107,000. Maximum = midpoint x (1 + spread percentage / 2) = $145,000 x 1.35 = approximately $196,000.
Step 4: Round to a reasonable increment. Salary bands rounded to the nearest $1,000 or $5,000 are easier to communicate and administer than bands with precise dollar figures that imply a level of market precision the data does not actually support.
Step 5: Document everything. Record the survey sources used, the blend weights, the percentile target, the range spread rationale, the evaluation score that determined the grade, and the date the band was set. This documentation is your defense when the band is questioned, which it will be.
When to Update Salary Bands
Most organizations review and update salary bands annually, typically aligned with their compensation planning cycle. The review should examine whether market data has shifted significantly enough to require a band adjustment.
Triggers for an off-cycle band review include: significant market movement in a job family (a new wave of AI-related roles pushing technology compensation up rapidly, for example), a pay equity audit that reveals systemic compression within a grade, or the addition of new roles that do not fit existing band structures.
When you update a band, you do not automatically update employee salaries. A band adjustment changes the target range. Employee salary adjustments to bring people within the new range are a separate compensation decision that should be addressed in the next cycle or through an equity adjustment program. Use compa-ratio analysis to identify employees who are most out of position relative to the updated band.
Pay Transparency and Band Documentation
With pay transparency laws now active in California, Colorado, New York, Illinois, Washington, and multiple other US states, salary ranges must be posted in job advertisements. A posted range that is not grounded in a documented evaluation and market analysis methodology is not just a compliance checkbox. It is a potential litigation exhibit.
When you post a salary range, you should be able to answer the following questions in writing: How was the midpoint determined? What surveys were used? What percentile was targeted and why? What job evaluation produced the grade this range is tied to? If you cannot answer these questions in writing, your band is not defensible regardless of whether the numbers look reasonable.
CompBldr's Compensation Planning module maintains full version history for every salary band, including the survey data sources, blend weights, percentile target, and JESAP evaluation score that produced each grade placement. When the question comes, the answer is a thirty-second lookup.
Salary Bands Built on Three Inputs, Not Instinct
CompBldr connects job evaluation scores, blended survey data, and your pay positioning strategy to produce salary bands that are defensible from day one. Every band version-controlled. Every methodology documented. Every question answerable.
.svg)
.svg)
.svg)
