The Merit Matrix Chronicles: Setting Your Anchor

The second in a series (see first here) of posts - chronicles - on the ubiquitous merit matrix.

Many of my HR colleagues find designing and updating their merit matrix to be a daunting task.  All those boxes and numbers!  Where to begin?

In working with a matrix, I find it helpful to have a plan and my plan always starts with the first step of finding and setting my anchor.  The anchor is the box in the matrix where the "center" of the employee distribution lies; the bump in your bell shaped curve (or, in most cases, semi-bell shaped curve).

Let's say, for example, that the majority of your employees are in the middle third of their current assigned salary ranges and have had their performance assessed as "meeting expectations".  In this situation, and using the example matrix below, the anchor would be the cell indicated in light blue.  Typically, this anchor cell is where you place the salary increase percent that represents your overall budget, which (hopefully) bears some relationship to the average expected salary increase out in the marketplace.  For the sake of our example, let's say that this is 4%.

Once I have set my anchor, which gives me some perspective and a bit of framework from which to proceed,  I can move on to the other boxes in the matrix and determine the appropriate increase percents for each of them.

Of course, this first example is relatively straightforward.  A matrix which is anchored in the middle presents you with a less complicated design process.  What if, alternatively, you discover that your anchor is in a different place than the middle of the matrix.  What if most of your employees are in the lower third of their respective salary ranges and nearly everybody has had their performance assessed as being "exceed expectations"?  In this case, your anchor is the cell indicated in the second example matrix, shown below.

If the expected market increase level, and your salary increase budget, are 4%, then the anchor box in your matrix should be set at 4%.  Otherwise, your salary increase expenditures will be different (and in all probability exceed) your budget amount.  But this - and really any scenario where your anchor ends up "off center" - raises questions that must be addressed as part of matrix design process, such as:

• What is the most likely explanation for most of your employees being assessed as "exceeding expectations"?  Is your organization truly so fortunate as to employ so many individuals who have gone notedly "above and beyond", or (more typical) are managers simply taking the easy way out and giving everybody - whether deserved or not - a top performance rating?  If it is the former, you have a potential problem in that you can't reward all these employees at an "above average" increase level without pushing your entire budget up (which you might want to do in this scenario).  If it is the latter, then "exceeds" has become the new average performance level in your watered down merit pay system, which may have to be rewarded with an average level salary increase.  (Too bad for those employees who really did go "above and beyond".)
• Why are so many employees in the lower part of their salary range, below a competitive market pay level?  Is this due simply to a high volume of new hires, and a staffing strategy that involves bringing in relatively inexperienced people who will learn and grow on the job?  If this is the case, is it appropriate to deliver a market level increase to these employees, an increase that will not push them forward much in their range?  Or is this situation a result of chronically underpaying employees, in which case delivering a market level increase will have the net effect of holding them "in place" and furher exacerbating retention risks?

A matrix that is anchored somewhere other than the middle, as demonstrated above, demands a bit more consideration and strategy in order to ensure that you are spending salary increase dollars in the most fair and effective way.

Beginning the matrix design process by finding and setting your anchor, in my experience, gives you an initial stake in the ground from which to start, and helps you identify up front what challenges and issues must be addressed as you set salary increase guidelines.

Creative Commons photo "Sunset at Docklands" by Macinate

 

 

Comments

[Jill]

I'm curious how others determine what the remainder of the matrix should contain. Meaning, what is the tipping point for meaningful differentiation between rating and position in range? Myself, I typically start out by shear guestimate and see where that takes my budget. For example,if my anchor is set for achieveing - middle range is 4%, I may put 3% for achieving - high in range and 5% for achieving - low in range. I like to factor those by 150% for the high achievers, and give much less to minimal achievers. What'd everyone else's approach as a starting point to vet against budget?

[Ann Bares]

Jill:

Thanks for the comment.

You might find Hay's research on merit pay differentiation among Fortune's "most admired" companies to be of interest - see my post on it here http://compforce.typepad.com/compensation_force/2007/04/pay_differentia.html

Once I've set my anchor, my approach to completing the rest of the matrix is a somewhat "organic" one, moving through box by box. This is what I'd like to cover in my next "chronicles" post.

The process you mention also sounds like a reasonable one - sound guestimation with a careful eye on the budget.

[Frank]

Ann,

Can you please tell us why your merit planning does not start with the groups of employees that surveys show are the furthest behind their market comparators and perhaps need the largest pay increases. Isn't that a more logical starting point---taking care of the groups with the greatest needs first? Not sure myself which approach is best, just would like to hear your thoughts.

Perhaps you cover that in later posts on this topic.

Frank

[Ann Bares]

Frank:

I think of the merit increase matrix as being the general tool used for tying salary increases to employee performance, regardless of what position the employee holds. Jobs - or job groups - that are significantly behind market pay practices, to my mind, represent a separate pay issue than the design of the general merit matrix and are typically address through a separate process. Unless you are going to put in a different merit matrix for different job groups, depending on their competitive position - but I don't see a lot of that in practice.

Open to other perspectives and experience. Frank? Others?