Salary benchmarks are often used alongside a range of internal and external factors – such as affordability and other demands on the organisation’s resources, the pay bargaining environment (particularly where trade unions are recognised), the prevailing rate of inflation, and the fairness or otherwise of pay differentials within the organisation – to reach a decision about pay levels. You should start with understanding your organisation’s reward strategy, and the market you are hoping to compare against (this might vary per role, too). The data sources you choose must be credible and reliable, and you should understand where their data has come from.
You should refine the survey data to match the role that you wish to match, using seniority level, function, location and other filters to get a best fit. Following this initial look at the results, you may choose to set the filters wider, for example to take in a larger geographical area. You might need to rerun the data tables in parallel so that the varying influences of location, company size or industry can be assessed separately and a judgment made about the appropriate level of pay to be offered.
Further judgments are required to ensure that the “market” is defined widely enough. Employers often want to benchmark only against organisations that look like theirs. A more realistic approach is for you to benchmark against organisations that have employees that resemble yours. A key question to ask is, “What are the characteristics of the organisations from which we recruit staff and to which we lose staff?” Organisations that have those characteristics are the real market.