An Enduring Classic: The Prototypical Performance-to-Award Incentive Structure

When designing incentive plans and, particularly, when working to nail down plan mechanics, I like to trot out a little model that I position as the "classic" plan performance-to-award structure.  I was delighted to learn, in a seminar yesterday, that recent research conducted by the Hay Group (about which I hope to post in more detail soon) strongly validates the model I have been using for years, particularly for management incentive plans.

The model looks like this:

What does it mean?  I explain the terms and the model in this way:

Threshold means the point at which the plan begins to measure and award performance, the lowest level of performance that will earn an award.  In this classic model, Threshold is set at a performance level equal to 80% of Target (or the desired level of performance), and typically earns half the Target award level.  Say, for example, that we are looking at a very simplistic incentive plan where Target performance is simply $10 million in net income, and the participant has a Target incentive opportunity of 10% of their base salary.  Threshold, in this case, would be $8 million in net income, which would earn the participant an award equal to 5% of their base salary.

Maximum is the performance level which earns the highest possible award.  In this classic model, Maximum is set at a performance level equal to 120% of Target, and typically earns one and a half times the Target award level.

This is a model with leverage, as evidenced by the upswing and downswing in award levels relative to performance levels.

Should your incentive plan be designed exactly this way?  Maybe, but maybe not.  Some organizations don't pay any award until Target is reached, while others introduce some level of award at a performance level substantially lower than 80% of Target.  Some organizations prefer more upside leverage, and others prefer not to set a Maximum level at all.  Still others may find that it is helpful to identify five, or six, or more, levels of performance and award.

And - I have found - some performance measures (ratio-based measures, for example) do not calibrate well to this model at all.

What I find most useful about models like this one, is that they give us a starting point for design discussion.  Not a mandate to mimic, but rather a straw man to use as a point of departure.  It is with that spirit that I share this one here!

Comments

[Lisa]

Oh, I miss compensation plan philosphy and design discussions. With federally mandated pay scales, special salary surveys are about as much "open" discussion I am able to engage in - and even that is guided by a few too many rules :(

[Carla]

Ann,

Why doesn't the award increase in the same way as the target e.g., 110% of target equals 10 percent pay increase. Is there some research that backs your formula?

Thanks,

Carla

[Ann Bares]

Lisa:

I hear you! My sympathies for the "too many rules". May it not always be so.

Carla:

As I mentioned in the post, the difference in the "speed" of increase between performance and award is referred to as "leverage". Leverage is supposed to drive performance via increased upside for good performance (and, usually conversely, increased downside for performance below target). The formula I showcase here reflects "typical" leverage in non-sales incentive plans - which is backed up not only by my consulting experience and the experience of other compensation consultants I have worked with, but also the Hay research that I referenced in the post. But please note that the decision whether to use leverage, and how much leverage is appropriate, are not questions that research can answer for you - but rather judgment calls by management regarding the plan design that will yield the best outcomes, given a sound assessment of the situation and all the factors at play.