Chief Executive Boards Blog: Compensation Design -- 8 Steps to Pay (Only) for the Behaviors You Want

A few critical pitfalls pervade most small businesses. Lousy and late financial information is one of them. One of the others is poorly-designed incentive compensation programs. You'd be amazed at the millions of American workers who get incentive (or "bonus") checks for doing absolutely nothing above the minimum, or things that would have otherwise happened by themselves. And in some cases, for doing things that actually hurt the business, rather than help.

A member of Chief Executive Boards International brought this issue up in a Local Board meeting. He saw that the people selling his services were "farming" existing customers and not finding new ones. They were so comfortable with what they were making, they just didn't see a need to do the harder work of finding new customers. Yet he knew his business would die over time if this behavior continued.

Another member asked, "So, why don't you make it less comfortable to do what you don't want, and more attractive to do what you do want?" Good idea. Well, the member with the problem came back to his next CEBI meeting and reported that he had this problem fixed.

The solution? A slight twist to the compensation program. He set a quota of monthly sales to new customers. Sales people who didn't make their quota had their commission on all sales (existing and new) reduced by 20%. Those who made their quota got paid for both existing and new sales, and those who exceeded their new customer quota by a given amount got a 20% kicker on all sales (existing and new).

As we say in the South, "Nobody ever 'splained it to me that way before." Amazingly, everyone is making their new-customer sales quota now. Amazing -- get the rewards and penalties aligned with the business goals, and everything starts working better.

Here are 8 steps to designing an effective incentive compensation system:

Decide what "big result" you want -- sales, project outcome, etc.
Find a measure of that outcome that's closely related to the job -- something that the employee can actually control, and can see a way to impact. This is important -- plans based on "overall company profitability" rarely work -- the employees can't find the "lever" they're supposed to pull to affect a giant measure like that. Break it down into a measure they can understand and see how they can make a difference in.
Pick a "par" value for what "good" looks like -- a reachable level of performance that you'd be satisfied with on average. This should be something the typical person can accomplish -- not superstar territory.
Above "par", determine what you can pay for higher performance -- it should favor the "house". For example, if your gross margin on an additional revenue dollar is 40%, you might be willing to pay a sales person 10% (the house keeps 3 times what it pays out).
Consider "modifiers" that boost the employee's calculated bonus, based on other behaviors you want to reward. For example, new business vs. existing customers. Modifiers should be meaningful -- typically 10% to 50% of the base calculation.
Consider "demerits" that reduce the employee's calculated bonus, based on behaviors you want to eradicate -- for example, if sales reports aren't turned in on time or order entry information is incomplete. Or whatever else just "bugs you" about employee behavior. Again, meaningful percentages.
After you've designed (or redesigned) the plan, then turn your hat around and ask, "If I were the employee, how would I "game" this system?" Then go back to steps 5 and 6 and put in safeguards to prevent those abuses.
Lather, rinse, repeat -- if they think of games you didn't, adjust the plan as needed. This "modifier" idea is really powerful -- you can play with the modifiers at least annually. The base objective may not change, but you have the prerogative to tweak the modifiers to suit yourself or the company's current (and changing) strategies.

Important Note: Step #4 above is critical -- make the plan sensitive enough by starting at "par" rather than zero on the performance metric. Plans that pay from the very first unit of measure cannot possibly be sensitive enough to get an employee's attention -- it's simply not worth the effort to go the extra mile. Here's an article that describes that important strategy in detail: http://www.chiefexecutiveblog.com/2008/02/1-incentive-compensation-plan-design.html

If you've had particular successes or failures in compensation plan design, please click "Comments" below and share them with others.

Other CEBI Blog Articles...

To forward this to a friend, Click Here