The Ten Commandments of Performance Measurement

Employees who have a stake in the financial success of their organization are much more focused and engaged in their work and aware of how it impacts the overall success of that organization than those who don’t. Unfortunately, most organizations rely solely on increases in base pay to reward employees and struggle with how to get the most from their compensation dollars, including how to evaluate and compensate performance objectively and equitably. Great care should be taken in developing specific measures for scorecards and incentive pay. There are some measurement principles that apply to all measures and should be considered each time a measure is developed. The following is an excerpt from William Abernathy's book Pay for Profit.

1. No one should design their own incentive plan. The manager, or someone other than the individual receiving the incentive payments, should design the scorecard. The incentive plan designer can solicit input from those who will be assigned the scorecard, but should not negotiate the measures [SIC]. Otherwise, the participants will be put in a self-serving position which may prejudice the design. The employees in the job are often too close to the work and tend to design measures around processes and activities, rather than true results. In the design of incentive plans, management should think of itself as the "customer" and the participants as the "vendors." The customer always specifies what he wants and what he is willing to pay for it.
2. The frequency of measurement feedback is as important as the incentive amount. The more frequently the measures can be reported, the more effective the measures will be in guiding behavior. Try to implement measures that, at a minimum, provide monthly feedback.
3. Design ideal measures, then compromise. Determine the strategic results of a job position and the key performance dimensions (productivity, quality, sales, etc.). Design the scorecard to improve these results without regard to what data are available. Then capture the data or compromise. Don't design the scorecard exclusively to what data are currently available.
4. The performance measures should "mirror" the real world. Try to design scorecard measures as though the participants are franchised or in business for themselves.
5. Where possible, design measures for small teams and individuals rather than large groups. Individual measures have more impact on performance, are more equitable, and can more readily convert to leveraged incentive pay. Team measures may be combined with individual measures on the same scorecard, when appropriate.
6. Measure only controllable job outputs. Design measures that are largely under the control of the participants. Do not use broad financial or subjective measures affected by events the performer cannot control.
7. Balance quality and quantity. Never design one-dimensional scorecards that focus only on work quantity or quality. Make sure the two dimensions are balanced in terms of the economic consequences and the impact on long-term objectives.
8. Design "linked" measures to encourage employees in interdependent jobs to cooperate. When the performance of one employee group consistently affects another employee group, you can improve cooperation by including one or more scorecard measures from group A's plan in group B's.
9. Provide equity of opportunity, but not necessarily equity of result. All participants should have an equal opportunity to achieve the maximum incentive, but not necessarily every employee, nor every month. Employees should never come to see maximum achievement as guaranteed, or as an entitlement.
10. Try it, then fix it. All the variables that may affect performance will only surface after implementation. Scorecards should be piloted using non-monetary recognition, or low payout opportunity "capped" cash plans. Once you test the plans and make adjustments, increase or remove the incentive caps.