Measuring the Performance You Really Want
By Jone L. Pearce
Professor of Organization and Management
There is no doubt about the importance of accountability, but accountability for what, exactly?
The power of what is measured to drive performance is clear. The difficulty is not in understanding this but knowing how to set up job performance measures you really want. Perfect measurement of employee performance is very rare in practice. If it were possible, organizations could hire contractors rather than employees. This means the popular hope expressed in the box below is nearly always a false hope.
True or False?
We just need to design a better performance measurement system.
We have employees because we cannot precisely measure everything we might want from them in advance, and we will always depend on their willingness to exercise good judgment. But we still want employees to have clear expectations about what they should do and be accountable for their actions. This dilemma can be addressed in part by carefully avoiding the following impediments to measuring the employee performance we really want.
MEASURING QUALITY
The measurable actions and performance outcomes are not always the actions and performance outcomes most needed. A rigid adherence to what is most measurable can result in goal displacement: neglect of what you really want to achieve in favor of what is most easily measured. This can lead to serious problems for organizational performance.
One form of goal displacement is a sole focus on performance quantity because it is easily measured: sales volume, deadlines met, earnings targets exceeded, and the like. For most jobs, these outputs should be measured.
However, alone they are incomplete measures of most jobs because they say nothing about performance quality, something that usually requires someone’s subject judgment after the fact. Sales volume may be important, but not if it comes from front loading the orders and so robbing sales from the next reporting period. Deadlines must be met, but not by producing incomplete and thoughtless reports.
Furthermore, if good job performance is defined solely as numeric targets, no one should be surprised if employees produce quantity without helping customers and co-workers, problem-solving, and exercising good judgment. They are just doing what they have been told the organization really wants.
Goal displacement can be avoided by assessing performance quality and contextual performance along with quantity measures. For example, sales associates might also be judged on the number of calls to clients and client evaluations of the associates. Examples of good judgment and citizenship can be recorded and discussed during feedback sessions.
However, quality and behavioral assessments can rarely be specified with the same precision as performance quantity measures. Assessments of work quality and employee effort often rely on someone’s judgment, and so are subject to interpretation and dispute. Such measures will be contentious and never perfectly satisfactory to everyone. Recognizing the imperfection of job performance measurements is fundamental to managerial maturity.
MEASURING INNOVATION
Specific, difficult goals encourage employees to meet these goals, but because innovation is inherently unpredictable and unknowable, it cannot be subject to strict accountability for goals met. Such goal accountability can be interpreted as high evaluation pressure, leading to less experimentation and innovation. In practice, employees are admonished to innovate, but innovation is rarely measured.
Years ago a major bank established a program to promote innovations in new products and customer service. However, employees’ performance was still measured as meeting routine targets such as number of new customer accounts. There was little innovation. This bank’s attempt to encourage innovation is all too common: because innovation could not be defined in advance with specific goals and measures, it was left completely outside the performance management system. Organizations that say they want innovation, but hold employees accountable for what is most easily measured, will not get innovation from their employees.
Research indicates that having multiple performance measures, such as current job performance and innovation, does not impede experimentation as long as there is low evaluation pressure. Just because innovation must be assessed after the fact does not mean it cannot be assessed as part of expected job performance.
WHAT IS REALLY IMPORTANT WILL CHANGE
Change within organizations is inevitable. And change continuously undermines existing employee accountability systems.
All measurement systems decay over time, something that Marshall Meyer has called the performance paradox, showing that once a measure of performance is used to evaluate, those being evaluated will seek to improve their performance on that measure. Over time, they either learn how to perform better on that measure or are removed for poor performance and who are evaluated on that measure will do well on it. This inevitable process leads organizations to continuously add new measures that can better differentiate performance. But again, performance inevitably improves on the new measures, and so other measures should be added.
Performance paradox is why we often find very baroque measurement systems, as evaluators constantly seek to discover the better performers, and performers constantly seek to improve on their performance measures. Over time, measurement systems can become extremely elaborate and complex, with ever more time and resources devoted to maintaining them. The more performance measurement is used to differentiate employees, the more severe this decay will be.
Too often, this ratings inflation is not correctly attributed to the external cause of the performance paradox, but is falsely attributed to soft-hearted performance raters. This mistaken internal attribution has led to attempts to force dispersion (i.e., must have high and low performers in a predetermined distribution) on a measure on which all employees are actually performing identically. Such strongarmed human resources policies are rightly decried as arbitrary and unfair by both the employees and the managers rating their performance.
These forced performance distributions are all too frequent, and as Marshall Meyer has persuasively demonstrated, based on an ignorance of the inevitable decay of all measures used to differentiate performers. It is fairer and more credible to add new measures than to force a distribution of ratings on a performance measure with no actual differences between high and low performers.
Inevitably, attempts to change and adapt organizations are undermined by existing employee measurement systems. The more detailed the performance management system, and the more specific the goals, the greater the likelihood that existing accountability systems will undermine change. The more sophisticated and elaborate the measurement system, the more resistant everyone is to changing it.
How exactly will employees need to allocate their efforts on these new initiatives? Employees worry about how they will fare under new systems. If the new performance measurement system is not calibrated correctly, will they still be judged fairly? Furthermore, change requires managers and employees to devote considerable time to redesigning performance measurement systems, and this takes time away from other job responsibilities.
Tips on Performance Quality:
• Approach performance measurement as you would diagnose any quality improvement effort: use process flow charts and cause-and-effect diagrams to identify wasteful practices, rework, sources of conflict, and opportunities for improvement.
• Performance problems should first be viewed for any possible problems in system design rather than rushing to blame individuals.
• Initiative, quality and innovation should be identified as necessary components of job performance, and examples of successful performance should be identified and discussed during performance reviews.
• Employees should be held accountable for quality improvement suggestions as well as for quantitative performance goals.