Bottos, who is Vice President of Compensation at data provider Kenexa, offered her suggestions at a recent webinar sponsored by the company. Here are her 10 steps:
1. Analyze your organization’s jobs to determine:
- Main function of the job and key responsibilities
- Knowledge/skills/abilities required
- Organization level and reporting relationships
2. Develop a brief job summary:
- “Capsule” job descriptions facilitate survey matching
- Note that existing summaries may not be up to date
- Summaries are not a replacement for a full job description (I’ll bet 80% of your job descriptions are not up to date, Bossos says.)
There are three common problems with job summaries, Bottos says:
- The writer doesn’t know the job.
- The summary is woefully out of date.
- The summary is written for the person, not for the job. Think of the next person, not the incumbent, Bottos recommends.
3. Identify benchmark jobs:
- Common jobs within a recruiting market
- Typically found in many types of organizations (e.g., accountant)
- Can be industry based (e.g., nurse)
There can be non-benchmark jobs, Bottos notes. For example, new or evolving jobs or very specialized jobs. Most companies can match about 60-70 % of jobs. If you buy a lot of data, you can go higher.
Are class action lawyers peering at your pay practices? It’s likely, but you can keep them at bay by finding and eliminating any wage and hour violations yourself. Our editors recommend BLR‘s easy-to-use FLSA Wage & Hour Self-Audit Guide. Try it for 30 days … on us.
4. Select survey or surveys:
- Geographic (e.g., Boston Metro)
- Functionally-based (e.g., Accounting)
- Industry-based (e.g., Healthcare)
You might want to mix these, Bottos says. And don’t try the “We’re so unique, there is no data” approach on her. The fact is that you don’t pay people differently, so where are you getting talent from? And who are you losing it to?
5. Identify appropriate matches:
Consider the following elements, Bottos says:
- Are titles and levels within those titles consistent?
- Are JDs robust enough to use as a basis for a match?
- Do the descriptions provide clearly defined levels such as education level and years of experience?
In this market, says Bottos, be aware that some employees have been “paid” with new title. (It’s title inflation,” she says)
Again, don’t expect a 100 percent overlap on your matches, Bottos says. Look for about a 70% overlap.
6. Choose the appropriate market scope and determine market value at pay points corresponding with compensation philosophy:
- Geography, industry, size
- Mean (average)
- Median
- Quartile
Many target the 50th percentile, Bottos says, but she has one client who selects the 25th percentile. They don’t have much money, but they are fair across the organization.
On the flip side, another client targets the 90th percentile, but only for scientists. They aim to hire the best post-docs and work them hard.
There is no right or wrong in this; it’s about what works best and what you can afford. Also, you don’t need same reference market point for all employees.
7. Adjust data based on credibility of source and degree of match:
- Weigh survey data
- Assess premium or discount
If you need to adjust up or down more than 20 percent, you have the wrong benchmark, Bottos says.
All you need to avoid exempt/nonexempt classification and overtime errors, now in BLR‘s award-winning FLSA Wage & Hour Self-Audit Guide. Find out more.
8 Age the survey data based on the organization’s market timing and methodology:
Lead. To lead, if you are going to adjust on January 1, age it forward to January 1 of following year.
Lag. To lag, use this January 1.
Lead/lag. To lead/lag, age the January 1 data to June. You’ll lead for first half, lag second half.
Regardless of which approach you choose, do the same across all jobs, Bottos says.
9. Slot non-benchmark jobs based on comparable value to organization:
An organization job may fall between two survey jobs or require knowledge, skills abilities that are similar to another benchmark job. It’s OK to take part of one job and part of another,
10. Analyze variance versus current practices:
- What is the gap between how your organization currently pays and where it wants to be in the market?
- What action should you take with individual employees?
Challenges in Market Pricing
Bottos sees the following problems most often:
- Matching specific jobs to survey benchmarks
- Finding enough surveys for a particular market
- Benchmarking new jobs
- Going beyond data to make sure that your plans reflect your specific organization
In tomorrow’s Advisor, the Sherman Anti-Trust concerns of comparing notes about pay, and an introduction to a rather extraordinary FLSA self-audit guide that will help you find problems before the feds do.