• Startseite »
• Artikel »
• The Value of Pay Data on the Web

By John A.Menefee, Ph.D.,
Watson Wyatt Data Services
When it comes to obtaining online pay data, compensation professionals should not confuse access and convenience with accuracy and completeness.

QUICK LOOK

Compensation professionals need tools and processes to evaluate the reliability of online compensation data.

New, nontraditional Web sites offer data for many different reasons.

Some sites provide pay data to generate traffic for a wide array of products and services.

Few new, nontraditional sites conduct valid salary surveys to obtain data.

Compensation professionals face the daunting challenge of setting and implementing appropriate pay practices in a rapidly escalating war for talent. Hire the best, retain the brightest, but maintain a competitive and balanced pay structure. Such a task requires having current benchmark pay data, along with the compensation policies and practices that underlie current pay levels.

At first glance, some new, nontraditional salary survey sites appear to offer a seemingly efficient way to access the pay data needed to maintain a competitive pay strategy. This new way to acquire compensation information provides rapid access possibly to one or more data sources. However, how does pay data delivered over the Internet, in contrast to data delivered on diskette, CD-ROM, or in print, enhance its real value?

To answer this question, compensation professionals need some type of evaluation tool or process to help them make an informed decision about the role and value of online compensation data sites.
A simple checklist might help, focusing on the following:
The primary purpose of the data site and its host
The site's targeted customers
The underlying data base for the reported pay data
The methodology used to gather and analyze the data.

The following discussion, hopefully, will help develop and expand a more exhaustive evaluation checklist by providing some "rules of thumb" regarding the validity, reliability and usefulness of online data sites.

Who Is the Web Site Host?

From large search engines to trade associations, Web sites are offering some type of compensation data. As a first step in evaluating the usefulness of a data site, look to see who maintains the site. Site ownership can help clarify the real purpose or goal of the site. It's important to ask: What is the site's primary business?

The majority of the established compensation survey organizations currently host some type of Web site that provides either access to data or the opportunity to order data from their ongoing compensation surveys. These sites simply offer another distribution channel besides the more traditional methods of print and electronic media, such as CDs and diskettes.

In contrast, the new, nontraditional data providers offer only pay data over the Web for many different reasons. Several sites use the lure of compensation data to "bait and catch" individuals to build databases of self-reported data. Sites best described as virtual storefronts are providing access to pay data as nothing more than a traffic generator for the marketing of a wide array of HR related and non-HR related goods and services. Others are using pay data as a means to build a job applicant referral database for recruiters.

A number of sites that call themselves e-businesses are "resellers." They are middlemen or jobbers simply pulling together similar and disparate data sources selling the convenience of access to one or more data sources on a single or multiple job basis.

For example, some of these sites simply take data out of classified ads in newspapers, job postings in trade publications, or off other Web sites and resell it as current market data. Others are taking public access, government data files and are compiling databases that contain estimated data 3-to-4 years old. Because of the disparity in the makeup of these new sites, paying attention to the site host and its underlying purpose or motivation will help the user initially evaluate the value and applicability of the data such sites offer.

Who Is the Customer?

The second most important question to ask: Who is the site's ultimate customer – businesses (employers) or individuals (consumers)? Sites targeting individuals are more prone to offer the purchase of "one job at a time" salary data. These sites are interested in generating a high volume of transactions with little interest in building or maintaining any type of customer service or relationship.

Sites targeted toward businesses should provide more than just salary or total cash data. Supporting documentation related to the data survey source and job descriptions for the position data reported are a minimal requirement to accurately access the data's reliability, and the potential for any type of continual use.

Data Sources

Customers purchase name brand products because they know the use of highquality ingredients and proper processing/continuity results in a reliable, quality product. When purchasing a new or unfamiliar product, greater attention and scrutiny needs to be paid to the input sources and the processes used.
This is especially true when purchasing something like information on the Web where little, if anything, may be known about the inputs or the processes of the information provider.

Few, if any of the new, nontraditional sites offering pay data conduct statistically valid salary surveys to obtain their data. It is important to check the extent of documentation on the data sources used. Is the data self-reported by individuals, or is the data from employer-based surveys?

A description of the format, scope and purpose of the survey instrument used to gather the pay data and the data submission process should be available. Look for a list of the survey participants and a demographic profile that describes the number, size, location and industry composition of the survey participants.

What compensation elements were surveyed and how well do they fit with what you are paying? Are job descriptions provided for positional pay data and is there any information about the strength of the position match by the survey participants? Mismatched jobs and not properly accounting for pay by job level are the most common mistakes made in pay level comparisons.

It is extremely important to verify how the data has been acquired, especially for the new, nontraditional sites. There is a growing concern surrounding the resale of survey data by secondary parties. Recent court rulings on the application of copyright laws to electronic media will make it illegal to resell data without permission of the survey originator. This undoubtedly will cause several of the newer, nontraditional sites to either rethink or drop the offering of pay data as an end product or sales/marketing ploy. Reliance on these sites for pay data needs may be short-lived and problematic.

METHODOLOGY

Whenever survey data is used, close attention should be paid to how the data is analyzed. Look for answers to these basic questions.

What experience and expertise does the data provider have in survey methods and statistical analysis?

How has the data been screened or cleaned to eliminate mismatched jobs or erroneous values? One or two poor job matches can easily make the "average" pay level far from its true value.

Have multiple data sources been aggregated together to build a larger database? Aggregated or compiled databases may be mixing data that has values calculated with different statistical methods.

There are seven different ways to calculate an "average." Has the data been "aged"or updated? If so,was it done on a geographic, industry or revenue basis?

Finally, when was the data gathered? What is its effective date? Mismatched data, in terms of effective dates, can result in erroneous benchmark data for both salary planning and new hire pricing.

Data Output

The type of information provided over the Web typically is limited in its scope and depth. Most nontraditional sites offer what is called "one job (pay level) at a time." For $29 to $225, one can purchase the average pay of a selected position. It is quick and relatively cheap. The difficulty comes in interpreting just one data point. For example, $23,500 is given as the average salary for a benefit analyst. Is this the total annual amount of direct cash compensation? What about bonuses, profit sharing, or in today's high-tech world, broad-based stock options? Knowing the pay policies and practices of the companies who are the pay data providers also is important, not only to properly interpret average salary levels, but also to know how the salary component fits into total pay.

Reliability is Critical

For most companies, compensation is the single largest annual expenditure. To remain competitive in the war to attract and retain talent, compensation data has to be reliable, supportable and complete both for salary planning and market pricing. The availability of pay data on the Web by a growing number of new, nontraditional data providers raises questions and concerns about their motives, processes and products.

ABOUT THE AUTHOR

John A. Menefee, Ph.D., is managing consultant at Watson Wyatt Data Services, a global organization that provides compensation surveys using information compiled from a database of more than 750,000 incumbents.

FOOT NOTES
Visit our Web site at www.worldatwork.org and
go to Information Central.There you will find
ResourcePRO, a powerful database that holds
almost 4,000 full-text documents on total
rewards topics.
For more information related to this article:
. Log in to ResourcePRO and select
Simple Search
. Select a Rewards Category: Compensation
. Type in this key word string on the search
line: "market data"OR "methodology and
salary" OR "internet and salary"