How to Create Salary Bands from Scratch (With a Free BLS Data Walkthrough)

Why Most HR Teams Build Salary Bands Backward — and How to Fix It

Picture this: you're the HR manager at a 75-person logistics company, and your CEO has just asked why two account managers at the same level are $22,000 apart in pay. You open the spreadsheet that passes for your compensation structure. There are no defined ranges — just a column of salaries added one hire at a time over six years. You have no documented rationale, no benchmark anchor, and with pay-transparency laws now covering a significant and growing share of U.S. employers, no defensible answer for the next job posting.

The honest truth is that most growing companies skip formal salary bands until a moment exactly like this one forces the issue. And when HR finally sits down to build them, the temptation is to work backward from what you're already paying — which just entrenches whatever compression, inconsistency, or inequity already exists in the data.

The better starting point is external: publicly available wage data from the U.S. Bureau of Labor Statistics, specifically the Occupational Employment and Wage Statistics (OEWS) survey. It's free, methodologically rigorous, and updated annually. This walkthrough will take you from that raw data to a working set of pay bands — including the formulas, the decision points, and the validation checks — without requiring a compensation consultant or a five-figure software platform.

By the end, you'll know exactly how to create salary bands that are defensible, market-aligned, and ready to attach to your job levels.

Step 1: Understand What a Salary Band Actually Is

Before the math, a quick grounding — because the terminology varies widely and getting it wrong costs you later.

A salary band (also called a pay band or salary range) defines the minimum, midpoint, and maximum cash compensation the organization will pay for a given job level or job family. The midpoint represents your market target — typically the wage percentile you have decided to compete at. The minimum and maximum express how far you're willing to pay above and below that target while keeping someone in the same band.

Three terms matter for the mechanics you'll use below:

• Midpoint — your competitive anchor; the dollar figure you benchmark against market data.

• Band spread (or range width) — the percentage distance from minimum to maximum, expressed as (max − min) ÷ min. A spread of 50% is common for professional individual-contributor roles; narrower spreads (30–40%) suit more structured hourly or entry-level jobs; wider spreads (60–80%) suit senior leadership roles with longer development arcs.

• Compa-ratio — an employee's salary divided by their band's midpoint, expressed as a percentage. A compa-ratio of 1.00 (100%) means the employee is paid exactly at market midpoint.

If you want a deeper treatment of the underlying logic before diving into data, the complete guide to job band structure covers the full framework.

Step 2: Pull Your Benchmarks from BLS OEWS

The BLS OEWS program produces annual employment and wage estimates for approximately 830 occupations, covering the nation, all 50 states, and roughly 530 metropolitan and nonmetropolitan areas, drawn from a sample of about 1.1 million establishments. That breadth makes it the most practical free benchmark source available to HR teams at small and mid-size companies.

Here is how to use it:

1. Go to bls.gov/oes and open the "May 2025 National Industry-Specific and by Ownership" data or the most recent release. For most SMBs, the national occupational estimates table is the right starting point; you can layer in state or metro data once you have a baseline.

2. Match your jobs to SOC codes. Every OEWS estimate is organized by Standard Occupational Classification (SOC) code. You need to identify which SOC code corresponds to each role you're building a band for. This is not always obvious — a "Logistics Coordinator" at one company might map to SOC 13-1081 (Logisticians) or to 43-5061 (Production, Planning, and Expediting Clerks) depending on actual duties. Match on duties, not job titles. The guide to matching jobs to SOC codes walks through the decision logic in detail.

3. Read the percentile columns. For each SOC code, OEWS publishes the 10th, 25th, 50th (median), 75th, and 90th percentile annual wages, plus the mean. Use the 50th percentile (median) as your default midpoint anchor if you're targeting a market-median pay position (the most common choice for companies without an explicit competitive pay philosophy). If your organization has decided to pay at the 75th percentile to attract talent in a tight market, use that column instead.

4. Check the geography. National figures are a reasonable starting point, but wages for the same occupation can vary substantially by metropolitan area. If your workforce is concentrated in a high-cost metro — say, the Denver-Aurora-Lakewood MSA in Colorado or the Seattle-Tacoma-Bellevue MSA in Washington — pull the state or metro-area file for a more accurate anchor. The BLS OEWS benchmarking guide covers how to navigate the geographic files.

For a Canadian workforce, Statistics Canada's Table 14-10-0417-01 provides NOC-based wage percentiles by province and census metropolitan area under an open government licence, and follows a similar percentile structure to OEWS. Specific figures should be drawn directly from that table for the relevant NOC code and geography.

The all-occupation national mean annual wage in the May 2025 OEWS release was $69,770. Use occupation-specific figures — not this mean — for your band midpoints; the all-occupation mean is a useful sanity-check anchor only.

Step 3: Set Your Competitive Pay Position

Your midpoint is not simply whatever BLS says the median is. It is whatever BLS says the median (or your chosen percentile) is, adjusted for the pay position your organization has consciously decided to compete at.

Before you calculate a single band, answer these two questions in writing:

What percentile do we want our midpoints to represent? Common choices:

• 50th percentile — market median. You pay what the market pays; your value proposition to candidates includes factors beyond cash (mission, flexibility, career path, benefits).

• 75th percentile — above-market. You are competing primarily on cash. This approach is sustainable for critical roles in tight labor markets; it's expensive to apply universally.

• A blend — some organizations pay the 50th percentile for most roles but the 75th for a handful of hard-to-fill technical positions.

How will we handle geographic variation? If you have employees in multiple states or metros, you need a consistent policy: one national midpoint for all, location-adjusted midpoints by tier, or full metro-level differentiation. There is no universally right answer; the right answer is the one you can administer and explain.

Document both answers. They are the foundation of your pay philosophy, and they determine which column you pull from OEWS. Changing them later without documentation is how pay inequity re-enters the system quietly.

Step 4: Calculate Min, Midpoint, and Max

With a midpoint figure in hand, the band math is straightforward. The formulas are:

Min = Midpoint ÷ (1 + (Spread ÷ 2))
Max = Midpoint × (1 + (Spread ÷ 2))

Or, for a commonly used approximation:

Min = Midpoint × (1 − Spread ÷ 2)
Max = Midpoint × (1 + Spread ÷ 2)

Both give you a band that is symmetric around the midpoint. The first formula is mathematically precise; the second is easier to audit in a spreadsheet. For spread values up to about 60%, the two produce nearly identical results — the difference only becomes material at very wide spreads.

Worked example. Suppose you are building a band for a mid-level Logistics Coordinator role and you have pulled the 50th-percentile OEWS annual wage for SOC 13-1081 (Logisticians) for your state. For illustration purposes only, substitute your actual OEWS figure: assume the median is $82,000. You have decided on a 50% spread for this level.

Min = $82,000 ÷ (1 + 0.25) = $82,000 ÷ 1.25 = $65,600
Max = $82,000 × 1.25 = $102,500

So your band runs $65,600 → $82,000 → $102,500, with the midpoint at the market median. An employee at the midpoint has a compa-ratio of 1.00. An employee at $75,000 has a compa-ratio of 0.91 (91%) — below midpoint, consistent with being newer to the role or earlier in development.

The min/mid/max band math explained article covers the algebra in more detail, including how to back-calculate a midpoint from an existing salary distribution when you are retrofitting bands onto current pay.

Step 5: Choose the Right Spread for Each Level

Not every level in your structure should use the same spread. A wider spread creates more room for pay progression within a band before an employee needs to be promoted; a narrower spread pushes people toward the top of a band more quickly, creating pressure for promotion or reclassification.

General guidance by level type:

Level typeTypical spread rangeWhyEntry-level / hourly30–40%Shorter skill development arc; promotion points are closer togetherProfessional individual contributor40–60%Standard range for most SMB professional rolesSenior IC / specialist50–70%Longer mastery arc; meaningful performance differentiationManager / director60–80%Variable scope, team size, and business impact justify wider rangeExecutive80%+ (often variable-comp blend)Broad authority; often supplemented by bonus or equity

These are informed starting points, not fixed rules. Your spreads should reflect the actual career arc of each level in your organization and the degree of performance differentiation you want to make visible in base pay. For more on this calibration, see the band spread and range width guide.

Step 6: Build Multiple Levels into a Coherent Structure

A single salary band is a data point. A salary band structure is what gives employees and managers a legible career framework and gives your organization the ability to do pay-equity analysis across levels.

When you lay multiple bands side by side, you need to decide how much they overlap. Band overlap is the percentage of one band's range that falls within the adjacent band's range. Some overlap is expected and healthy — it allows a highly experienced Level 2 employee to earn more than a newly promoted Level 3 employee before the Level 3 person gets up to speed. Too much overlap (more than 50–60%) collapses the distinction between levels and makes the structure hard to explain.

A practical check: for any two adjacent levels, the lower level's maximum should not exceed the higher level's midpoint by more than a small amount. If it does, either your spreads are too wide or your midpoints are too close together, and you should revisit one of them.

A five-level professional IC ladder might look roughly like this (all figures illustrative — replace midpoints with your OEWS-sourced anchors):

LevelMarket anchorSpreadMinMidpointMaxL125th pctile40%$42,000$50,000$60,000L250th pctile50%$56,000$68,000$85,000L350th pctile50%$72,000$88,000$110,000L475th pctile60%$88,000$112,000$142,000L575th pctile70%$104,000$135,000$176,000

This is a teaching example only. Use your actual OEWS figures for your specific SOC codes and geography. Do not apply these dollar amounts to any real compensation decision.

Notice that L2 max ($85,000) is well below L3 midpoint ($88,000) — the overlap is modest. L4 and L5 use the 75th-percentile anchor because these are senior roles where you have decided to pay above market.

For a complete walkthrough of how levels connect to titles, career frameworks, and leveling criteria, see what are job bands.

Step 7: Validate Your Bands Against Current Pay

Once you have a draft structure, map every current employee into it. For each person, calculate their compa-ratio: Salary ÷ Band Midpoint.

What you are looking for:

• Compa-ratios below 0.80 (below 80% of midpoint) — employees paid significantly below your market target. These represent equity risk, retention risk, and — in an increasing number of jurisdictions — pay-transparency compliance risk if a posting for their role would now disclose a range they sit below.

• Compa-ratios above 1.20 (above 120% of midpoint) — employees paid at or near the top of their band, often called "red-circled." They may be at career ceiling in their current level, or the band may need to be revised upward.

• Employees who fall entirely outside any band — typically a sign that the job was titled or leveled inconsistently when they were hired.

This validation step almost always surfaces surprises. That is expected. The goal at this stage is not to immediately adjust every salary — it is to understand the gap between where you are and where your structure says you should be, so you can prioritize remediation and plan a budget cycle to close it.

Step 8: Document Your Methodology — Before You Post a Single Job

This step is the one most teams skip, and it is the one that creates the most legal and operational exposure.

Your band-building methodology should be documented in enough detail that someone else on your team — or an external auditor — could reconstruct what you did and why. That documentation should cover, at minimum:

• The data source (BLS OEWS, specific release date, national vs. state/metro)

• The SOC code(s) used for each job family

• The competitive pay position chosen (which percentile and why)

• The spread methodology and any level-specific departures

• The date the structure was set and the planned review cycle (annually is standard)

Why does this matter before posting? Because in California, each non-compliant job posting is a separate violation — and "we hadn't finished documenting our ranges yet" is not a defense. With 17 states and multiple municipalities already enforcing active pay transparency laws affecting an estimated 65% of U.S. employers (Lift HCM, 2026), the window for treating compensation structure as a back-office project is closing fast.

Documentation is also what makes your salary bands a system rather than a one-time exercise. Without it, the bands erode within 18 months as managers negotiate exceptions and new hires are placed wherever the budget allows.

Putting It Together: When a Spreadsheet Stops Being Enough

The process above is genuinely doable in a well-structured spreadsheet — for a small number of bands, a single job family, or a first-pass structure to get your organization off a blank page.

But spreadsheets carry a structural limitation: about 94% of business spreadsheets used in decision-making contain errors, according to research cited by Phys.org and Frontiers of Computer Science (2024). When the thing being managed is compensation — with legal exposure attached to every posting — a formula error in your band-math tab is not a minor inconvenience.

As you scale beyond a handful of bands, or as you need to produce pay-transparency-ready ranges for every job posting, or as you want to run compa-ratio reports across the whole organization without manually updating a pivot table, you need a tool built for the purpose.

The Job Band Structure Builder is designed for exactly this transition point: it imports BLS OEWS anchors directly, calculates min/mid/max across multiple levels with configurable spreads, maps current employees to their bands, and exports the formatted ranges you need for compliant job postings — without requiring a compensation-consulting engagement. If you want to see how the methodology in this article translates into a working structure for your organization, the pricing page shows what the toolset costs.

What to Do Next

Building salary bands from scratch is a structured problem, and you now have the methodology:

1. Pull your occupation's wage percentiles from BLS OEWS (or Statistics Canada for Canadian roles), matched to the right SOC or NOC code and the right geography.

2. Choose your competitive pay position deliberately — and write it down.

3. Apply the min/mid/max formula with a spread appropriate to the level type.

4. Lay multiple bands side by side and check overlap.

5. Validate against current pay and flag compa-ratio outliers.

6. Document the methodology before the first posting goes live.

The hardest step is usually the first one: getting your job families matched to the right occupational codes so the benchmarks mean what you think they mean. The BLS OEWS benchmarking guide and the SOC code matching guide are the right next reads for that piece. The complete job band structure guide covers the full architecture — levels, titles, and career frameworks — once your math is solid. ```