How to Do a Compensation Analysis: Template, Steps, and Example

Key Takeaways

• A compensation analysis compares what you pay each role against the external market and against comparable employees inside the company, then turns the gaps into costed recommendations. It is a structured process with a handful of ratios at its core.
• The template that goes with this article has 22 columns and five formulas: compa-ratio, range penetration, market ratio, adjustment to minimum, and total cost. Those five numbers carry the whole analysis.
• Most analyses fail at the input. A market benchmark collected six to twelve months ago is already behind, and every ratio built on it inherits the lag.
• The output is a compensation analysis report a CFO can act on: which roles sit outside the range, what it costs to fix, and what happens if nothing changes.

A compensation analysis is a structured comparison of what an organization pays for each role against two references: the external market rate for that role, and the pay of employees doing comparable work inside the company. The output is a set of ratios for every employee or job, a list of the roles outside an acceptable range, and a costed plan to correct them.

The mechanics are not complicated. What separates an analysis that changes pay decisions from one that gets filed is the quality of the market number it is built on, and whether the results reach leadership in a form they can act on.

So let's take it in order: the template, the six steps, a worked example, and the report.

What is a compensation analysis?

In practice, it is two comparisons run on the same dataset.

• External, or a compensation market analysis. Each job is matched to a market benchmark for the same job, level, and location. The company's pay is expressed as a ratio to that benchmark. This answers whether you are competitive.
• Internal, or an internal equity analysis. Employees in the same job and level are compared with each other. Pay differences are tested against legitimate factors like tenure, performance, and location. This answers whether you are consistent and fair.

Why run both? Because they fail in opposite directions. A market compensation analysis alone will tell you a role is paid at market while two people in it are paid 30% apart.

An internal-only analysis will tell you everyone is consistent while the whole job family has drifted 15% below the market.

Scope changes the name but not the method. A job compensation analysis looks at a single role or job family across all incumbents.

An employee compensation analysis looks at every person in scope, which is what a full annual review does. The template below handles both.

The compensation analysis template

The template is one sheet with one row per employee. Make a copy of the Google Sheet, or rebuild it in your own spreadsheet from the columns below.

Inputs from your HRIS and payroll (one row per employee)

Inputs from your market data and salary structure

Calculated columns (the formulas)

Two columns at the end are filled by hand after the review: Recommended adjustment (currency) and Rationale (one line: market, internal equity, promotion, or no change). The total of the adjustment column is your budget ask.

Fill it in this order. Export the HRIS and payroll columns for everyone in scope. Paste the market median for each row from your benchmark source.

Paste the range minimum, midpoint, and maximum from your salary structure. The formulas do the rest.

If the job matching is already done, the whole file populates in an afternoon. If it takes a week, jobs are being matched to the market one at a time by hand, which is the step market pricing software automates.

Six steps from scope to sign-off

Here is how to conduct a compensation analysis, start to finish.

1. Set the scope and confirm the job architecture. Decide whether this is a full employee compensation analysis or a job compensation analysis for one family. Then confirm that every role in scope has a level and a title that maps to a market job.

Analyses go wrong at this step more than any other. If two people with the same title are doing different jobs, every ratio downstream is wrong for one of them. A usable job architecture is the fix.

2. Pull the internal data. Base, bonus target, equity value, level, location, tenure, and rating, straight from the systems of record. Do not accept a spreadsheet someone maintains by hand as the source; reconcile it to payroll first.

3. Match jobs to the market and pull the benchmarks. For each job and level, find the market median base and total compensation for the right location. This step decides the quality of the whole analysis.

A benchmark collected six to twelve months ago describes a market that no longer exists for fast-moving roles. Pave's compensation benchmarking data comes from integrated HRIS, payroll, and cap table systems across 8,000+ companies, so the median you paste in reflects what those companies pay today.

4. Run the ratios. With inputs in place, the five formulas populate. Sort by market ratio first to see external competitiveness. Then sort by compa-ratio to see structural position. Then filter the flag column.

5. Test internal equity. Group by job and level, and look at the spread of compa-ratios within each group. A spread wider than about 20 points needs an explanation in the tenure and rating columns. Where the explanation is not there, you have an internal equity issue.

If the analysis is also serving as a pay equity audit, add the protected characteristics as a separate column set and run the same comparison.

6. Cost the adjustments and write the report. Fill the recommended adjustment column, total it, and package the findings in the report format below.

Compensation analysis example

The fastest way to see the template work is one job at one level. Four professional-level software engineers in the same Tier 1 location, a range of $150,000 to $190,000 with a $170,000 midpoint, and a market P50 base of $175,000.

Employee A needs at least $8,000 to reach the range minimum and about $33,000 to reach market. Tenure and rating decide where between those two numbers the recommendation lands.

Employee D is above the range maximum. That opens a promotion conversation or a range-width question; it is never a pay cut. Employees B and C sit where the structure intends.

Now run the same four rows against a market P50 of $185,000 instead of $175,000, which is the difference a year of market movement can make for an in-demand role.

Employee B's market ratio drops to 0.89 and picks up a flag. That is the whole argument for the input quality in step three. The analysis is only as current as the market number in the column.

What goes in a compensation analysis report

The report is where the analysis earns its budget. Keep it to five sections.

1. Scope and method. Who was analyzed, which benchmark source and date, and which ranges were used.
2. Market position. Median market ratio by job family and level, with the families sitting below 0.90 called out.
3. Structural position. Distribution of compa-ratios and range penetration, plus the count of employees below the range minimum and above the maximum.
4. Internal equity findings. Job-and-level groups with unexplained spread, and any pay equity findings if that check was in scope.
5. Recommendations and cost. Adjustments by priority (below minimum first, then below market, then equity corrections), the total, and the projected market position after the adjustments.

A report structured this way answers the three questions a CFO will ask, in order: are we competitive, are we consistent, and what does it cost to fix. Anything else goes in an appendix.

Where compensation analyses go wrong

Three failure patterns account for most analyses that get filed instead of funded.

Stale market inputs. A compensation market analysis built on a survey collected last cycle compares current pay with last year's market. Every ratio inherits the lag. The report either understates the gap or, worse, calls a role "at market" when it is not. The fix is a benchmark that updates as the companies in it change pay, which is what payroll-integrated data provides.

Mismatched jobs. Matching a senior engineer to a mid-level benchmark because the titles look similar produces a flattering market ratio and a wrong recommendation. Job matching should be checked by someone who knows the roles, and ideally automated against a maintained job catalog.

Ratios without a decision rule. A compa-ratio of 0.87 means nothing until you have decided that 0.85 is the floor and anything below it gets adjusted next cycle. Set the thresholds before running the analysis and write them into the flag formula. The report then writes itself.

Run the analysis on real-time data

The template works with any market source. What changes with Pave is the number in the Market P50 columns and how long it takes to get there.

Pave's salary benchmarking data comes from integrated HRIS, payroll, and cap table systems, so the median reflects current pay instead of a survey cycle. Market pricing automates the job matching, applies geographic pay differentials, and keeps every range in one place instead of a spreadsheet with formula errors.

Request a demo to run this analysis on your own roles against real-time benchmarks.