Reflecting on performance appraisals…
Thirty years of research into performance appraisal consistently shows that our approach consumes much time and produces little or no return.
William Edward Deming, the father of quality management, listed ‘evaluation of performance, merit rating and annual review’ as the third of his ‘seven deadly diseases’. These are the ‘diseases’ Deming considered the most serious barriers management faces to improving effectiveness and continual improvement.
Those participating in an annual performance appraisal might long to be at Adobe. It has been reported that Adobe has banished the formal yearly performance appraisal in favour of a far more sensible approach. It is an approach that positions assessing performance as a relationship between grown-ups rather than a tick-a-box process that adds little value to the lives of managers or employees.
Performance appraisal is a surprisingly recent but amazingly persistent management fad that remains in most organisations in nearly the same form as it originated.
We begin by reviewing the preceding year, we set goals for the coming year, we have a mid-point ‘check-up’, and we close the year by looking back at what was completed. We might ‘rack-and-stack’ those at the same level across the company to make additional judgements about ‘talent’, and then we recognise and reward as necessary. So, what’s the problem with the way performance appraisal is done?
The problem is that thirty years of research into performance appraisal consistently shows that our usual approach consumes much time and produces little or no return. For instance, research regularly finds that most people need more guidance on improving their performance due to their performance appraisal. Reviews of the literature show that regardless of the stated purpose of the performance appraisal system—feedback, development, organisational alignment, pay—there seems to be limited evidence that it is a process with any positive effects. And we have known this for a long time.
For example, in 1986, a study across 84 different performance management systems in 35 different organisations found that:
The majority of people who are rated less than the highest on a rating scale disagree with the rating.
The majority of people who disagree with the rating are less motivated and less dissatisfied with their jobs after the appraisal.
The majority of these people reported having ‘little or no idea’ how to improve their performance.
The differences between self- and supervisory appraisals are largely a function of perceptions that there were factors beyond the employee's control influencing the outcome.
Those being rated focus on external factors, while those doing the rating focus on the attributes of the individual.
That performance appraisal has become such a fixed part of business practice is difficult to fathom.
Changing the terminology from ‘performance appraisal’ to ‘performance management’ in the late 1970s to bring in a more comprehensive approach to individual assessment has not improved the situation. The attempt to humanise and broaden the performance appraisal system likely made it more complicated.
The media reporting Adobe’s decision to abolish annual performance reviews has conveyed awe and wonder. A moment to reflect on our own experiences of the process (as a manager and employee) and the considerable research that shows it delivers limited value suggests that Adobe was making a sensible business decision to stop wasting resources. They didn't stop investing in assessing and improving performance; they decided to do it differently. Not brave, just sensible.
Various organisations survived and prospered without formal performance appraisal systems for a long time. This raises the question: why do formal performance appraisal systems exist? I guess a question like this set off the chain of events that led to Adobe’s final decision. Maybe, more leaders should be asking this question, or better still, what is our current approach to performance management contributing to organisational capability?
