.svg)
What is a composite in compensation?
A composite is the result of blending pay data from multiple market sources into a single reference value for a role. Instead of taking the P50 from one survey and using it as the market reference, a compensation team using a composite approach pulls the relevant percentiles from several sources, applies weights to each based on their relevance and reliability, and combines them into one blended value.
The composite is typically the last step in the market pricing calculation before the market reference point (MRP) is established. It represents the organization's best single estimate of what the market pays for a given role, level, and location, based on all available data rather than any one source.
Why composites produce better market references
No single compensation data source is complete. Different surveys cover different industries, geographies, and organizations. Some have large sample sizes for certain roles and thin data for others. Some publish annually; others publish more frequently. Each has its own methodology for collecting and reporting pay data.
Using a composite addresses these limitations in several ways:
• Reduces dependence on any single source: if one survey has an outlier value for a role due to sampling variation or methodology, the composite dampens its effect rather than letting it drive the entire market reference.
• Improves coverage: different sources may have stronger coverage for different job families or industries. A composite that draws on multiple sources is more likely to reflect the full range of organizations competing for the same talent.
• Increases defensibility: a market reference derived from multiple sources is easier to defend to leadership than one based on a single dataset. It signals that the compensation team did not simply pick the most convenient number.
• Enables source comparison: building a composite requires the team to look at what each source says for a given role. When sources agree, confidence in the reference is high. When they diverge significantly, that is a signal to investigate the match quality or consider whether the sources are drawing from different talent pools.
How composites are built
The composite process typically follows these steps:
• Identify relevant sources: select the data sources that have reliable coverage for the role, level, and market cut being priced. Not every source needs to be included for every role.
• Extract percentile values: from each source, pull the relevant percentiles (P25, P50, P75) for the defined market cut, including geography, industry, and company size parameters.
• Apply aging: bring each source's data forward to a common effective date using an aging factor, so all sources reflect the same point in time before being blended.
• Assign weights: decide how much weight each source should receive in the blend. Weights are typically based on sample size, relevance to the organization's industry and talent market, and the compensation team's confidence in the source's methodology.
• Calculate the weighted average: multiply each source's value by its weight and sum the results. The output is the composite value for that percentile and role.
The resulting composite P50, P75, and other percentiles become the market reference points used in pay range construction, offer decisions, and benchmarking analysis.
How weighting works in a composite
Weighting is the most judgment-intensive part of building a composite. There is no universal formula for how to weight sources. Common approaches include:
• Equal weighting: each source receives the same weight. Simple to explain and apply, but does not account for differences in sample size or relevance.
• Sample-size weighting: sources with larger sample sizes for the relevant role and market receive proportionally more weight. More statistically rigorous but requires access to sample size data from each source.
• Relevance weighting: sources more closely aligned with the organization's industry, size, or talent market receive higher weights regardless of sample size. Requires judgment about which sources best represent the actual competition for talent.
• Hybrid weighting: a combination of the above, applying higher weights to sources that are both large and relevant. The most common approach in practice.
Whatever weighting methodology is used, it should be documented and applied consistently across pricing cycles. Changing weights without documentation makes it difficult to understand year-over-year movement in market references and creates audit risk.
Composite vs. single-source pricing
Some organizations price roles using a single data source rather than building a composite. This is simpler and faster, and may be appropriate when only one source has reliable coverage for a given role or market. The tradeoff is greater sensitivity to that source's specific methodology, sampling, and any year-over-year changes in its data.
As a general principle, the more strategically important the role and the more the pay decision matters to retention and competitiveness, the more value there is in building a composite. For high-volume, lower-stakes roles where multiple good sources exist, a composite produces a more reliable and defensible reference. For highly specialized roles where data is thin across the board, a single well-matched source may be the only practical option.
What is a composite in compensation?
A composite is a blended market reference point created by combining pay data from two or more external sources into a single weighted value. Rather than relying on one data source, a composite draws on multiple sources to produce a more robust and representative pay benchmark. It is a standard output of the market pricing process and the basis for most market reference points and pay ranges.
Why use a composite instead of a single data source?
A composite reduces dependence on any single source's methodology or sampling variation, improves coverage across different industries and job families, and produces a more defensible pay reference. When multiple sources agree on a value, confidence is high. When they diverge, that signals a problem worth investigating before the reference is used in pay decisions.
How are sources weighted in a composite?
Common weighting approaches include equal weighting (same weight for each source), sample-size weighting (larger samples receive more weight), relevance weighting (sources more aligned with the organization's industry and talent market receive more weight), and hybrid approaches combining several factors. Whatever method is used should be documented and applied consistently across pricing cycles.
How many sources should be included in a composite?
Most organizations use two to four sources for a composite. Using more sources is not always better, it increases complexity and can dilute the contribution of the most relevant sources. The goal is to include sources that genuinely improve coverage or reduce reliance on any single dataset, not to maximize the number of inputs for its own sake.
What happens if sources in a composite disagree significantly?
Significant divergence between sources in a composite is a signal worth investigating before the composite is finalized. Common causes include different job matching (the sources may not be pricing the same role), different market cut definitions, or different publication timing. Understanding why sources diverge is as important as knowing what the final composite value is.