RSS icon

Top Stories

How to Conduct an Effective Pay Survey

April 7, 2002
Related Topics: Compensation Design and Communication, Featured Article, Benefits

One of the hardest jobs in managing employees is making sure they’re happy.Although there are many aspects to creating happy employees, a fair andcompetitive wage is still the number one yardstick in measuring satisfaction.After all, how many people do you know that work only because they love the workthey do?

How do you make sure you are paying your employees fairly? The answer’senough to fill a textbook. You should always be "selling your story" to youremployees not only in terms of pay, but also work environment, benefits,prospects for advancement, and the other so-called fringe benefits. But thefirst place that dissatisfied employees look is the marketplace, and if you wantto stay ahead of the curve, you should look there too.

First, a word about collecting market pay data. There are so many ways tocompare your pay against the market -- published pay surveys from consultingfirms, online Web survey companies, association groups, and even the Departmentof Labor. For common, homogeneous jobs, these surveys work great. But forspecialized jobs, especially managerial and above, large amounts of reliabledata are hard to come by. That’s why there’s no more certain way to collectthe right data than through a custom survey. It’s not difficult to do as longas you have your ducks in a row.

Pros and Cons
Customized surveys have several advantages:

  • They can be targeted directly at the companies with the closest match foryour position.

  • You can collect current salary data, rather than data that was collected12 months ago.

  • You can specify exactly the information you want to collect, rather thanporing over general salary surveys.

They also have some distinct disadvantages:

  • They can be costly to administer in terms of time and money. However,those costs need to be compared to purchasing survey services, which can beexpensive also.

  • They can come under more legal scrutiny with regard to pay discrimination(i.e., several companies collude on wage rates, or "price fixing").

  • Developing a good survey instrument is challenging.

Once you’ve determined the information you want to collect, organizing itin an orderly fashion is not too hard. Some of the most typical pay surveyquestions ask for the following information:

  • Minimum, midpoint, and maximums of a company’s pay range for theposition.

  • Current rate of pay.

  • Number of people in the position (that will help you find the average).

  • Average length of service of a person in the position.

  • Availability and amount of bonus payments.

How to create a good survey instrument
Most custom surveys start with these basic elements:

  • Job Title: sounds simple, but as you’ll see in a minute, it’s veryimportant.

  • Benchmark Job Duty Features: The most important element. These are theessential job duties that, along with the job title, will guide your surveyparticipants in selecting the position within their company to report.Generally, keep the list short; only three to four maximum. As a rule of thumb,use the duties that take up at least 50 percent of their time. You don’t wantto include duties such as planning the annual company picnic (unless it’s anessential duty).

  • Match Strength: For each position, include an indicator that therespondent can use to gauge how close of a match their position is to yours --generally a one to ten scale. If it’s a low match, such as a one or two, youmay want to exclude the data. If it’s higher, but not strong (five or six) youcan adjust the data in relation to strength of match.

  • Pay Data: As mentioned above, you want to gather enough information toanswer your questions, without asking for too much. Not all companies have thesame compensation system, so you’ll want to keep your data fields as generalas possible.

The overriding theme of a good survey is "simple." If you want goodresponses, you need to make it very straightforward for the respondent and notoverly burdensome. Therefore, don’t send out a survey for 20 positions, askfor eight different pay sources, and ask for it back in a week. Compensationanalysts will turn their nose up at you every time.

The survey process
As it goes with most things, garbage in, garbage out. The root of any goodcustom survey is a well-designed job description. That sounds logical enough,but it’s surprising how many companies have not done a good job analysis. Ifyou have put together a tight job description, then picking the benchmarkfeatures should be easy. Again, you want to make it simple for the HR person onthe other end to find the match in their organization.

Once you’re ready to send out your survey, there are a few time-honoredtips to keep in mind:

  • If your survey is somewhat involved, pretest it on a colleague orcoworker. Unclear syntax or directions can be cleaned up before "going live."

  • Call candidate respondents ahead of time to get their OKs onparticipating. You may need to play salesperson a little. Be upfront about thedetails of the survey and what will be expected.

  • Be reasonable -- but firm -- on response time. Never use phrases such as"at your earliest convenience" or "ASAP," but state clearly the responsedate. Keep in mind that certain times of the year are worse than others,particularly in the Fall when budgeting and pay adjustments for many companiesare being planned.

  • Have multiple options available for respondents to give you information.Depending on the size of the survey, a simple phone call will do. Some are finewith paper surveys, but many might like to get a spreadsheet and to send it backvia e-mail (NOTE: the convenience of spreadsheets sent via e-mail is attractive,but take some time to work out security measures, such as encryption and virusprotection, so that respondents are comfortable with receiving and sending paydata).

  • Check in a couple of times with respondents to make sure the survey isunderstandable and also to remind them of the survey deadline.

The other important thing to do, even before you decide you need to survey,is develop a network of survey respondents. Local chapters of human resourcesorganizations, such as SHRM, are good places to start. If it is difficult tofind respondents in your local area, look at cities within a reasonable distancefrom your location and realize that you may need to adjust your data forcost-of-living. If certain individual respondents are hard to come by, thenperhaps industry associations will have pay survey data that can be used.

The results are in
Now that you’ve received your data, it’s time to analyze it and matchagainst your data. Allow enough time in your survey process to be able to goback to your respondents to clarify any confusing information that they sent.Then get to work on compiling the information so that you can report back toyour respondents with the summary data.

You should take no longer than two weeks to finish up your analysis of thesurvey data and report back to respondents.

As with anything done well, organization is key to a successful effort.Whether short or long, you can conduct a successful survey as long as you giveyourself and your respondents enough time.

Workforce Online, April 2002 -- Register Now!

Recent Articles by William Dickmeyer, CEBS

Comments powered by Disqus

Hr Jobs

View All Job Listings