I posted a few months back about the indefatigable compa-ratio, a simple statistic that can provide a wealth of information about your base salary practices.
A number of readers commented about related approaches and tools; I thought a follow-up post might be in order to address some of the great questions they raised. Snooze alert for those of you who can't bear much compensation detail; here is a boatload comin' at you.
Salary Range Quartiles
Alan mentions his organization's use of quartiles to manage base salaries. Sovan asks the question: "I am comfortable with compa-ratio but not with converting the same to quartile or vice versa. Can you suggest an article/post which can be helpful in studying this?"
Dividing a salary range into quartiles is another time-honored method for managing base salary practices, and is an approach often featured in conjunction with a merit increase matrix, as Alan points out. I urge my clients to define their quartiles as target salary areas for different employee groups; e.g., the first quartile (lowest 1/4 of the salary range) represents the appropriate salary for new, relatively untried and inexperienced employees.
As a salary management tool, quartiles (or quintiles, or deciles, or thirds/tri-tiles) are less precise than the compa-ratio statistic (as they describe a "chunk" or "portion" of the range, rather than a specific dollar point), but they can serve an important purpose nonetheless.
In terms of Sovan's question about conversion, I typically define range quartiles using compa-ratio. For example, a salary range that is 50% wide from minimum to maximum, with a midpoint exactly in the middle, will have a minimum equal to 80% of the midpoint (the same as a compa-ratio of 80%) and a maximum equal to 120% of the midpoint (the same as a compa-ratio of 120%). You could further use compa-ratio to define quartiles in this particular salary range as follows (not in a perfect geometric sense, but in a simple and straightforward one):
Quartile 1: Compa-ratio of 80% to 90%
Quartile 2: Compa-ratio of 90% to 100%
Quartile 3: Compa-ratio of 100% to 110%
Quartile 4: Compa-ratio of 110% to 120%
Salary Range Percentiles
Lupe says: "At work we use SAP Compensation system which uses compa-ratio, but management uses percentile. As an analyst I always have a problem with translating percentile to compa-ratio. Do you have a formula for this?"
If I am correctly understanding the use of the term percentile here, I believe Lupe is referring to an alternative statistic to the compa-ratio. A client of mine uses a percentile statistic in lieu of compa-ratio which he calls "point in range", and which he believes is a more intuitive figure for communicating with managers and their employees. Having never heard of it before, I assumed he invented it (Dave??), but perhaps it is in more widespread use than I realized. At any rate, point in range is calculated using this formula:
(employee salary - range minimum)/(range maximum - range minimum)
And it works this way:
If an employee's salary is at minimum, the "point in range" is 0% (that is to say, he/she has not yet made any progress into the range)
If an employee's salary is at midpoint, the "point in range" is 50% (which indicates that he/she is exactly halfway through the range)
If an employee's salary is at maximum, the "point in range" is 100% (which indicates that he/she is at the top - or highest level - of the range)
The exact translation of compa-ratio to point in range will, unfortunately, vary in accordance with the width of the range - except at midpoint where point in range will be exactly half of compa-ratio.
That's it. Thanks to all who commented on the compa-ratio post for your willingness to take a deep dive into the quantitative end of the salary management pool!
Follow-up note: Point in range, or range penetration formula corrected above - thanks, Dan for noting the error.
Image: Jane M. Sawyer
I believe the "point in range" formula is
(employee salary - range minimum)/(range maximum - range minimum)
We refer to it as "range penetration." It shows the relative pay compared to the range, and we use it where I work to (A) keep people from focusing/obsessing on the midpoint, and as you discussed, (B) use portions of the range as target levels for certain categories.
It might be interesting to see if others believe the focus simply moves from "I need to be at the target/midpoint" to "I need to get all the way to the 100% mark."
Posted by: Dan Barber | May 28, 2008 at 03:44 PM
Dan:
Thanks for calling out the correction (this is what I get for composing posts late into the evening) - how great to have my compensation savvy readers backing me up.
Excellent point about the focus shift which may occur when using one statistic over the other: Moving away from "I need to be at midpoint" = good. Replacing it with "I need to be at maximum" = not so good. It may indeed be exchanging one communication challenge for another.
Any other readers with experience in one versus the other (compa-ratio versus point in range or range penetration)?
Posted by: Ann Bares | May 28, 2008 at 10:07 PM