The other day I was speaking with a sales compensation director who had observed a peculiar correlation in her sales comp data. She compared the payouts that salespeople received for their quantitative bonuses—measures of direct sales results—to the discretionary payouts those same individuals received. Like many asset managers, this company uses a framework for determining discretionary pay but provides significant latitude to sales managers in deciding the actual rating for each individual.


What she found was not unusual. In fact, I’ve spoken with two other firms just this month that have observed this exact same phenomenon, but I think it’s worth discussing. Here’s a masked summary of what she saw in her compensation data:

What she observed was a negative correlation between the manager-directed discretionary bonuses and the quantitative sales bonuses (that is, salespeople who earned more results-based pay earned lower discretionary pay, and vice versa). This correlation was significant and persistent as she found similar results in prior periods.


Why was this happening? Sales leadership was most likely using discretionary pay to buffer the impact of the quantitative sales plan. Sometimes this was done to correct perceived unfairness: for example, data that couldn’t be trusted or a short-term result that wasn’t reflective of the salesperson’s true contributions. Less often, it was done explicitly to ensure that a certain pay level was reached to “keep the person whole” according to their earnings expectations. And in many cases it may not have been a conscious decision at all but just a bias toward giving extra consideration to those who seemed to be viewed unfavorably by “the numbers.”


I told the compensation director that this was a classic issue with discretionary pay, but not the kind of “classic” you want to preserve. My take is that there are three things she could do, depending on the main source of the problem:

  1. If discretionary pay is plugging a hole in the performance measure, then fix the measure. It may be that the quantitative results really aren’t capturing the contributions of individual salespeople. If that’s the case, then look for more representative measures, or consider changing the amount of pay tied to the existing measures.
  2. If discretionary pay is plugging a hole in pay distribution, then reconsider the pay distribution. Perhaps the sales organization (or leadership) is more risk averse than originally thought, or perhaps payouts swing further period-to-period than they should. That could mean that the incentive plan needs to be recalibrated (longer horizons, lower volatility), or it could mean the sales team needs to change.
  3. If discretionary pay is being driven by an unconscious bias toward leveling, then address it openly and consider restructuring the calibration process. Sometimes simply creating awareness is enough to correct any underlying problems. In other instances, firms might benefit from rethinking how discretionary pay is determined: for example, moving it off-cycle from the quantitative pay process, or adding more governance.

Negative correlation between discretionary and quantitative pay should be an outlier in your compensation program. It really shouldn’t happen unless you have a small sample size, and it only happens in a single year (that is, it’s a true anomaly). But over long periods of time, we would expect salespeople who approach the business the right way to be those who generate the best quantitative results. If that’s not what you are seeing, then you might want to ask why.