Have you ever read about a company getting rid of performance management ratings and wondered, “How do they identify and reward high performing employees without some method to rate people based on performance?”
There are three methods that companies are using to resolve this dilemma. These are methods I’ve encountered through my work helping companies use technology to rethink performance management, which occasionally includes creating processes that eliminate manager ratings.
But before we get to that, I want to clear the air on some things about performance ratings in general.
1. No company is actually getting rid of ratings
The act of “rating” involves placing employees in categories based on their value to the organization. Any time a company puts one group of employees in a category that is perceived to be more beneficial than another, it has rated them.
Every company rates its employees whether they admit it or not. When a company pays some employees more than others it is rating them. If one employee gets a job assignment or promotion that others wanted but did not get, then these employees have been rated.
The question is not whether a company should rate its employees, because it will. The question to be answered is how can a company rate its employees in a manner that is accurate, efficient, effective, and done in a way that increases employee productivity and engagement?
2. Manager ratings of performance can be valuable
Eliminating manager ratings makes sense for some companies. But companies don’t have to remove manager ratings to have a good performance management process.
As seen in a recent performance management showcase, many companies get significant value from manager ratings. And sweeping criticisms of manager ratings commonly made in the press do not align with my personal experience working with companies or my professional interpretation of scientific data studying performance management.
For example, a recent article in Harvard Business Review, titled Most HR Data is Bad Data, stated:
… A significant body of research has demonstrated that each of us is a disturbingly unreliable rater of other people’s performance … no amount of training seems able to lessen it. And it is large — on average, 61 percent of my rating of you is a reflection of me … this effect has been well documented … in 1998 in Personnel Psychology; in the Journal of Applied Psychology in 2000; and in 2010, again in Personnel Psychology.”
I subscribe to Personnel Psychology and the Journal of Applied Psychology. The HBR article did not list the studies by name, but the only ones that make sense are Hoffman et al. (2010), Scullen et al. (2000), and Mount et al. (1998). I was struck by several things when I re-read these studies, particularly Hoffman et al., which is the most recent and most analytically rigorous.
First, the studies did not analyze performance management data. They analyzed data from 360 development surveys containing more than 50 individual ratings on specific behaviors.
The ratings were almost never tied to compensation or promotion decisions, were usually used solely for developmental feedback, and did not incorporate calibration methods commonly used to increase accuracy of performance management ratings. It is questionable how well research using 360 survey ratings generalizes to performance management ratings given how different they are in design and use.
Second, while the studies found that performance ratings were influenced by the unique perspective of raters, the researchers noted that this does not imply “biased or otherwise ‘incorrect’ ratings” and “these results reinforce the value of collecting performance data from raters occupying different organizational levels” (Hoffman, 2010, p.146 & p. 119, emphasis added).
Lastly, none of the studies investigated the impact of training on rating accuracy. Other studies have shown that accuracy of manager ratings can be improved through training (Roch et al., 2012).
People often interpret research differently, but from my perspective, the claims in the HBR article about accuracy of performance ratings are not supported by the research. Poorly designed performance rating processes will generate inaccurate data and performance ratings are undoubtedly influenced by things like manager personality.
But this does not mean that manager ratings are necessarily inaccurate or “bad.” For more discussion about differing interpretations of performance rating science, I encourage reading articles by Mark Effron, Susan Galer, and myself.
How to drop manager ratings and still identify high performers
Manager ratings can be valuable if collected as part of an appropriately designed performance management process. But for some companies collecting manager ratings creates more problems than it solves.
This is often true in companies that have a history of using rating methods that were highly inaccurate, controversial, or cumbersome (for example, companies that over-emphasized use of forced ranked ratings). For these companies, it can be more effective to eliminate manager ratings entirely instead of trying to fix the ratings.
The following are three (3) ways companies have eliminated manager ratings yet retained the ability to accurately measure, identify, and reward employee performance.
- Shift ratings from the employee assessment to manager calibration sessions. Traditional performance reviews involve having a manager and employee sit down on an annual or quarterly basis to review what the employee has accomplished, agree on an overall evaluation of the employee’s performance, and discuss future plans for employee development. Many companies are eliminating the second part of this process from the review. Managers and employees only discuss what the employee has accomplished and what to focus on going forward. There is no discussion or rating of overall performance. Instead, the information on employee accomplishments is used during subsequent talent review meetings where the manager meets with other managers to rate which employees in the company are contributing the most to the organization. This approach removes the anxiety and distraction created by including an overall manager rating in the performance review conversation. It also creates an incentive for employees to accurately document their accomplishments since this information will be used in subsequent talent reviews to determine their overall performance level. But these reviews do not happen until after employees and their managers have met to discuss future development plans.
- Evaluate employees entirely based on goal accomplishment. This process starts with managers working with employees to establish and maintain clear job goals. Employees meet with their manager on a regular basis to discuss, clarify, and if needed modify goals throughout the year. These ongoing discussions focus on improving goal clarity, supporting goal accomplishment, and evaluating whether individual goals, tasks, and projects are on track. No discussion is made about overall employee performance. Managers then reference employees’ goal accomplishments when making subsequent decisions about pay or overall performance. As one company using this process put it, “we focus on what people have done for the company and not who they are as individuals.” Companies that use this approach often include safe guards to ensure employee behaviors support company values lest people be rewarded for accomplishing the right things the wrong way. Some companies also calibrate goals based on difficulty and importance so certain goals are more valuable than others for enabling career advancement or compensation increases.
- Rate employees based on future actions instead of past performance. In this approach, managers are not asked to rate employees on performance, but instead rate employees based on recommended future actions that are typically associated with performance. For example, rating employees on things like “Do you think this employee should receive more or less pay relative to their peers,” “Do you feel this employee should be promoted over the next year,” or “How big an impact would the loss of this employee have on business operations.” These ratings are not necessarily shown to employees. The advantage of this approach is it skips the whole issue of measuring performance and focuses specifically on what actions to take with employees. The problem is it obscures the reasons why managers decide to favor some employees over others. This could create more inconsistency around how employees are treated and could make it hard for a company to understand and explain why rewards were given to some people but not others.
It’s time for change, but make sure it’s the right change
Most companies’ existing performance rating processes could be significantly improved, many are terrible, and some should be completely eliminated. But care must be taken when changing ratings that impact employee pay, promotions, and other career opportunities.
It is critical to think through the implications that will result lest you find yourself in the following situation (from a personal communication from a senior HR executive at a Fortune 1000 company):
We did away with ratings a few years ago, but it didn’t really work. The performance management system kind of collapsed as a result of taking away the operating mechanism. And now we are looking to revert back to the more traditional system.”
The challenge of performance management is effectively dealing with the reality that all employees are valuable but some employees are more valuable than others. This requires making ratings in a manner that accurately assesses people’s performance without damaging employee-manager relationships or making employees feel like they’ve been divided up into winners and losers.
Removing manager ratings may be part of this answer for some companies, But, success is usually more about what you create rather than what you eliminate.
