Salary surveys are our connection to the outside world. Just as scientists need to observe the world before drawing conclusions, so too do compensation professionals need to have data in order to understand the labor market. Good operating practices exist for dealing with market data to take full advantage of organizations' substantial investment in salary surveys.
For those who have been in compensation for many years, there has been a noticeable consolidation of salary surveys vendors. The industry now consists of fewer reliable providers. With quality data more scarce, prices have increased. In response, compensation professionals must strive for a balance of effectiveness and efficiency and be constantly attentive to price changes for surveys they regularly purchase. They also need to negotiate updated contracts with the provider and ask questions about whether survey packages can be customized to potentially save on costs. Organizations can also benefit from an on/off purchase strategy — buying a survey every other year and aging the data during off years.
The compensation function should be gathering as much data from reputable sources as possible — generally at least three that have good coverage of an industry or geography, which means spending more time participating in salary surveys to receive potential discounts, networking and asking peers in other companies which data providers they use. They also need to stay aware of the market data industry through product demonstrations and sales calls.
It is the compensation function's responsibility to be good stewards of the organization's investment in salary surveys. Therefore, it is important to always have an idea of what products are out there. By gathering more data, you'll be equipped to establish whether a particular survey is consistently higher or lower than another for similarly defined jobs. Despite a constantly changing workplace where jobs rapidly change or don't exist at all, survey data continues to be a mainstay for effectively managing compensation.
Although it is intuitive, averaging different surveys' results for similar jobs into a single number is a mistake for compensation specialists. Because vendors cannot share the underlying details of their surveys, you can't find out whether two surveys collected the same data in the same way from the same survey participants during the same participation period. Therefore, you can't say with any accuracy how you should weight the results in the calculation. Instead of making arbitrary rules for putting the results together, a better approach is to take multiple perspectives on the same benchmark job and phrase the result as a target number or range, assuming both results are close to each other and reasonable from a validity standpoint.
Another approach is to use the survey's capability to slice and dice the numbers by various cuts, if that capability exists. The most well-used data cuts in salary surveys are by revenue, full-time employees, industry and geography. As surveys have consolidated and sample sizes have increased, so too has the reliability of these data cuts; each has more incumbents on which to report. Leading survey vendors also provide the ability to create custom peer groups of five or more participants' compensation. With all of these resources at their disposal, compensation professionals should not be pressured to find and aggregate results from multiple surveys. Instead, it may be sufficient in cases where a job is rare and can only be benchmarked to one survey to only gather results from multiple cuts of that one survey.
Remember 60%/70%/80%. Fundamental to any survey use is proper job matching to the provided positions or levels. In general, organizations should aim to match at least 70% of their jobs to the positions in their survey libraries based on job content, according to blogger Ann Bares. Not all jobs can be matched to the survey, so it is acceptable to use the other jobs that are benchmarked to assign a relative value to those that are not. Using job content first and title only as confirmation, make sure the benchmark position from the survey captures at least 80% of the job content. If it does, then it's a match. Some jobs in an organization are so-called "hybrid" jobs; while combining results from multiple survey sources may cause an issue, combining results from the same survey source is necessary. In those cases, make sure the benchmarks accurately reflect the time spent in each area of responsibility. A good baseline is 60%, meaning the vast majority of jobs in an organization will spend at least half of the time working on responsibilities laid out in a survey benchmark.
Once the benchmark matches are solid, then compensation specialists can explore the possibility of sharing access to market data with HR practitioners in a self-service environment. Assuming there is a fundamental shared understanding of how salary surveys work, compensation functions can dramatically increase their productivity and accessibility by opening up their data through technology. Of course, this also depends on the kinds of systems in place in order for this kind of model to be implemented. Sharing data with HR generalists opens the door for organizations that want to be more transparent about pay with frontline managers in environments where the company culture supports it. And by having solid benchmark matches in place, there is little room for the HR community and managers to dispute what is being provided to them.
Compensation professionals need to make sure the organizations they support are getting the most out of their investment in market data. Good operating practices for doing so are to gather as much as they can, keep what they do gather separate and find the best matches possible before sharing with managers.
About the Author
Brock Meyer, CCP, GRP is compensation analyst for Valspar in Minneapolis.