Compensation 101: Selecting Market Data Sources

December 8, 2017

This is an excerpt from our new ebook “Comp 101: A Beginner’s Guide to Compensation Management.” Download the full guide here.

Since there are a variety of compensation data sources available to you, it’s crucial that you choose wisely. The two key things to consider when selecting your market data sources are:

Are they current, accurate and validated?
Do they cover the data needs of your jobs and organization, such as the breadth of your jobs, geographies and industries?

Types of Data Sources

The types of sources are varied and vast. Keep the above questions in mind as you explore.

Traditional (Standard) Surveys and Industry Surveys

This category includes some names you’re probably already familiar with: Mercer, Radford, Aon-Hewitt and Willis Towers Watson. Organizations participate in these surveys by matching their employees to survey job titles and descriptions, then submitting the data to the consulting firm. The firm then verifies and crunches the data to provide distributions back to the participants, and sometimes nonparticipants, typically for a fee.

Many firms offer industry-specific surveys that service only a given vertical. Trade associations will often make data available to their member bases as well.

Benefits: The methodology of this data source is well-understood. They typically provide a participant list, which shows you the businesses who participated in the survey. Usually this list contains mostly larger companies.
Disadvantages: Sometimes the data are broad or perhaps don’t provide info for more rural areas. And, because the data often lack freshness, they may have gaps for new and hot jobs.
Freshness: These sources are usually published annually — based on data that are up to nine months old. As a result, they often come with “aging coefficients” to be applied to the data.

Pre-Mixed Data

This category involves the most black-box methodology of all the data sources. It’s clear that the data are market data, but you can’t always be sure what the set consists of. The data provided by pre-mixed data providers are often a curated mix of traditional and industry surveys with some data modeling to fill in the gaps. An example of a pre-mixed data provider is Salary.com.

Benefits: Because of the mix of data sources and methodologies, pre-mixed data sets can often be one-stop-shops for compensation data. They can cover many of the gaps that show up in traditional data.
Disadvantages: There is low transparency in terms of data source with pre-mixed data. As a result, explaining either methodology or data to anxious managers may be a challenge.
Freshness: Because of the black-box nature of these sources, it’s unknown how often the sources are refreshed and updated versus aged or otherwise manipulated.

HRIS or Internal Data

One source of salary information is your own HRIS or other internal data. These data can be used to ensure internal equity. Extracting information from your own workforce can be meaningful.

Benefits: Internal data are great for looking at pay fairness. You can analyze them for compression, below-range pay and compliance concerns like gender pay equity and pay for other protected classes. This data set is an easy resource for running reports to compare departments or people within the same title.
Disadvantages: Gathering, sorting and structuring data can be a more manual process. Often analytics are difficult or have to be developed internally. And obviously, internal data do not give you visibility into the external market.
Freshness: The freshness will depend on how updated you keep your internal systems.

Crowdsourced Data

Crowdsourced data sets are what they sound like: data sourced from a crowd. In this case, the “crowd” is employees. Crowdsourced data providers, like PayScale, use a real-time survey as the data collection mechanism. When using crowdsourced data, it’s important to know the validation process for the data.

Benefits: Because employees know the most about their own jobs, crowdsourced data allow for much more specific and granular data. They typically cover more jobs and locations, as well as fast-moving and newly emerged jobs.
Disadvantages: Some groups are underrepresented with crowdsourced data. Often there is no motivation for executives to fill out online surveys; they know what they’re worth. Similarly, people in minimum wage jobs are less likely to fill out online surveys. Finally, online surveys tend to skew white collar, since they require easy access to a computer or smartphone.
Freshness: Crowdsourced data are updated on a daily basis. Things change in real time.

Crowdsourced data typically cover more jobs and locations, as well as fast-moving jobs.

Click to tweet

“Scraped” Data

These data are gathered from job listings. The practice of posting compensation data in job listings is much less common in the U.S. (about 25 percent of listings include salary) than in other parts of the world (greater than 80 percent include salary). Data can be extracted from listings using technology or browsing what’s available online.

Benefits: This is the only data source that provides direct insight into the demand for labor, since postings are created when a position is open.
Disadvantages: Scraped data can be messy, requiring a lot of moving pieces to gather the data. They also don’t reflect actual pay, since it’s not possible to infer from a posting what the incumbents actually received when they accepted the job.
Freshness: The freshness of scraped data is variable, since it requires the listings to be available with compensation data associated.

Government Data

The U.S. government provides some very broad compensation data trends. They’re available for some locations and some industries. They’re often fairly dated to the point that they’re no longer relevant or remotely competitive, but they’re free.

Match to Find Your Job in Your Data Source

The definition of a benchmark job is that it is a job that regularly exists in the market. (To compare your jobs with the market, those jobs need to exist in the market.) Here’s a helpful checklist of things that matter when comparing your job to the data source:

What is the essence of the job (not just the title)?
What level is the job (entry, professional, manager, director … )?
What are the top three job responsibilities?
What top three special skills are needed for the job?
Does this job have supervisory responsibilities?
What experience is necessary to do this role?

You may have some positions that you can’t find in your data source. Don’t force matches where they don’t exist; having no data is better than using bad data. You can decide how to pay the job by aligning the pay to a similar job inside your organization.

Before you move on to the next step …

Lock down your compensation data sources. Talk with data providers to get quotes. Ask them about their validation process and the depth and breadth of their data coverage. If you have critical roles to fill, talk with a data provider that has data that move more quickly.

The next step in the process is setting ranges — coming tomorrow. For all the steps and takeaways, grab your copy of the full Comp 101 guide today!