Calculated Benchmarks

What are calculated benchmarks and how are they computed?

• Calculated benchmarks are values that are computed using a combination of Pave's real-time data and pay differentials.

• We analyze median salaries for identical job roles at the same level to create a geographic pay differential. After comparing these differences, we apply rigorous statistical methods to construct calculated benchmarks that allow us to provide accurate compensation values across different geographical locations.

• Raw benchmarks, which are also available in Pave, use compensation information pulled directly from the HRIS to calculate the various percentiles available within the application.

• For every slice, we compute the calculated benchmark and compare that against the raw benchmark. We display whichever benchmark we have more confidence in unless we have Very High Confidence, High Confidence, or are Confident in the raw benchmark (has a sample size ~100+ and the margin of error is ~within 3 percentage points – see Confidence Scale) in which case we will default to displaying the raw benchmark instead of the calculated benchmark.

  • Confidence Scale:

    • Market data results are bucketed into the following Confidence Labels:

• There are two ways that we delineate calculated benchmarks from raw benchmarks in the product. At the top right of the page, there is an info icon indicating that the slice has a Pay differential applied. Additionally, in the table view, calculated values are underlined to help you quickly recognize them. 

• You can click the "Pay differential applied" icon to view more details on these calculations.