Performance evaluations do not work

A crazy thing is going on in corporate America. Ninety-five percent of Fortune 1000 companies use performance evaluations. That’s a whole lot of companies spending a whole lot of time and money to evaluate the performance of a whole lot of people. Yet research shows that, on average, performance evaluation programs have only a moderately positive effect on employee behavior and often hurt performance.

The popularity of performance appraisals goes back to a guy named R.B. Ammon. He published a summary of research on the subject in 1956. His analysis showed that feedback motivated people and “almost universally” improves performance. Ammon’s work was cited in more than 100 articles. In the academic world, that’s like having a top 10 hit.

Starting in 1985, researchers began to show that Ammon reached his conclusion by ignoring studies that showed mixed or negative effects from feedback. It was a bit of an academic scandal, but by then, corporations were under the spell of a performance review juggernaut.

Then in 1996, researchers Kluger and NeNisi analyzed as many well-designed feedback studies as they could get their hands on. They found that 38 percent of the time, feedback had a negative effect on performance. Assuming that in some instances the impact is neutral, that means that the odds of improving performance by using formal performance reviews are not much better than flipping a coin.

Why is performance feedback failing to do the job? It’s because the performance evaluation process is counter to fundamental motivation science — stuff that B.F. Skinner’s rats taught us years ago. Skinner studied the effect of reinforcement, that is, feedback, on behavior. He developed a rat feeding system that gave the critters a food pellet when they stepped on a bar. Performance reviews aren’t, and can’t be, designed to capture the motivational power of the mammal brain.

We use a memory aid called PRISMTM to help our clients teach Skinner’s insights on effective feedback to supervisors. PRISM stands for Precise, Random, Immediate, Sincere and Meaningful. Here’s how it shapes up:

Precise – Skinner didn’t reward his rats for being nice guys. He rewarded them for stepping on a bar. “Great job!” tells a person nothing about what they did right. It might make the person feel better, but it doesn’t make them do better. You have to tell the person precisely what they did right (or wrong) to improve performance. Make it about the work, not about them.

Random – Skinner experimented with different reward schedules. He didn’t give his rats food every time they stepped on the bar. Skinner discovered that an “intermittent reinforcement schedule” was the most powerful way to instill and maintain behaviors. If you reward on a random basis, people, just like rats, will exhibit the reinforced behavior more often than if you reinforced the behavior every time it pops up.

Immediate – When Skinner rewarded his rats with food, he didn’t wait until next week. He gave them their reward immediately after they stepped on the bar. Think of how performance reviews work – reinforcement comes annually, semi-annually, or maybe quarterly if the supervisor is really into it. That delay kills the connection between the behavior and the reward. Skinner’s rats would have starved.

Sincere – His rats didn’t seem to care if Skinner was sincere when he fed them, but people can read insincerity in the voice, the face, and the posture of a boss in the blink of an eye. For reinforcement to work with people, you have to mean what you say.

Meaningful – Skinner didn’t feed his rats sawdust, he gave them food pellets. Rats like food pellets and don’t care much for sawdust. The same thing goes for the rewards you use with employees. The reward has to be something they value. That doesn’t mean money. For a number of reasons, money is not the best reward. Intrinsic rewards have a stronger and more lasting effect. We think well-delivered praise is the most effective reward. It’s cheap, it’s easy, and it requires no paperwork.

Enough with the rats. Let’s apply PRISM to a workplace situation. You have a manager named Max who loses his temper (we’ll make Max a man, since men have more testosterone) when he works directly with a very personable and quick-witted salesperson named Minnie (we’ll make Minnie a woman, since women have higher oxytocin levels and a bigger corpus callosum). Max likes to ask lots of questions to really understand a situation. When dealing with Minnie, he sometimes loses his cool because, after answering, she quickly counters with clever questions of her own — questions Max struggles to answer. He gets flustered, raises his voice and the meeting ends with both parties on a high boil.

How are you going to fix this? You have two choices. You can go positive or negative, but you shouldn’t do both at the same time. Once you go negative, Max’s limbic system will take over and no amount of positive stroking will pull him back from his amygdala-made fight-or-flight response. He’ll just sit there stewing in his own juices.

Let’s look at negative first. You might talk to him on Monday, saying, “Max, I don’t want you losing your cool with Minnie anymore. You’re a manager by God, so act like one.” We’ve just violated a whole lot of Skinner’s teachings. The feedback is imprecise; what does “losing your cool” mean? And you’ve made it about Max instead of his behavior. The feedback is also delayed; you waited through the weekend.

You reconsider and talk to Max by the end of the day. “Please don’t raise your voice to Minnie. When she asks questions you don’t have answers for, just say, ‘I don’t know,” and go on with the conversation.” That’s better. But to make the lesson stick, you may have to reinforce that message from time to time, which is likely to irritate poor Max.

Let’s try a positive approach. Max works well with Leslie, another salesperson. Right after Max meets with Leslie, you say to Max, “I like the way you deal with Leslie’s questions. When he asks something, you say, ‘Tell me what you think.’ Could you try that with Minnie?” This way, Max doesn’t stew. To follow up, from time to time, but not every time he does it, you acknowledge Max for using the tell-me-what-you-think approach with Minnie and Leslie. That reinforces the behavior and soon you have Max showing interest in Minnie’s opinions.

Dump your annual performance review process. It’s not worth the time and effort your managers have to give it. Instead, teach your people what Skinner’s rats taught him. Be precise, random, immediate, sincere, and meaningful when you give feedback.

Bradley Harris is co-owner of NeuroSense Consulting, an HR consulting company specializing in neuroscience-based business coaching, management training and leadership development. He can be contacted at


Use the comment form below to begin a discussion about this content.

Sign in to comment