​JS:  I agree with all the reasons outlined but here’s what I would argue.  In some cultures, it does make sense to do an overall rating.  If there is a reward-oriented culture, you can make distinctions in performance, and if there is a significant amount of money on the table, it can work.
​

SSB: While many organizations are moving away from the rank and yank process (e.g., Microsoft) there are new adopters like Yahoo.  Are there better ways of addressing poor performers

​JS: I don’t know anybody who likes the rank and yank process. Here’s what happens when you tell people “you have to have someone at the bottom category at the end of the year and we will probably move that person out of the business.” They deliberately keep the weak performers around until the end of the appraisal period, so they have somebody to put in that box. What they ought to do is move the person out of business much earlier. Rank and yank encourages a lot of game playing and creates a lot of dysfunction.

​SSB: Given the inherent problems with the rating process, are we ignoring human tendencies that make such ratings flawed?  Studies show that we humans are not perfect, rational decision-makers and our decisions tend to be influenced by all kinds of biases.  Are we asking managers to do something that they are not humanly capable of doing? 

​JS: Of course my answer to that is yes.  But everything that managers do is subject to biases, and memory and processing shortcomings, and judgmental problems.  It’s not unique to just the performance appraisal process.

SSB: That’s probably the heart of the problem.  Any other replacement to the current system is likely to be equally flawed as long as manager judgments are involved.   This is probably a great place to implement big-data solutions in HR.  If we had more objective data about performance collected over a period, it would lead to more accurate ratings instead of managers relying on their memory.

​JS: I think that’s a noble idea, but I don’t think it’s going to solve the problem.  First of all, we got to make better use of data.  Nobody is going to argue about that.  But for many aspects of work performance, no objective data are available.  Even when we have objective data, that data tends to be deficient or contaminated.  Plus when people figure out the metrics, they try to drive their performance based on the data being captured.  That creates a whole bunch of distortions.  Another option is to get more opinions from many people.  I applaud that idea but that data is also flawed and subject to the same kind of systematic biases.
​
SSB:  Given all the challenges we’ve talked about, is there a way to have objective and fair appraisals?

​JS: Including more opinions than just the boss’s help.  It helps to have multiple sources of objective data, but even all that put together is an incomplete picture but better than a subjective opinion.  The problem is when we think the key thing is to do annual ratings well.  I don’t think that’s the key thing.

High-performing organizations realize that the key thing is realizing what’s happening day in and day out.  It’s the informal conversations and ongoing feedback that are important.  We know when you recognize people for doing good work they do more of it. 
Studies show that when we set challenging goals and people accept those goals, and we give them feedback along the way, and we praise them when they attain those goals- those things engage people, and we get higher levels of performance.  These are the things we should be focusing on.  These are very well-established psychological principles that have been around for a very long time. ​

The trick is to quit asking the question “how do we make better ratings?”  Accept that we can only do so much, there will always be a problem.  Whether or not you have an overall rating probably ought to depend on the culture of the organization and the goals of the organization.   What’s most important is to deemphasize the appraisal and emphasize what happens every single day. A performance appraisal is an annual event.  It happens once, and it’s a 30-minute conversation.  What about the 51 weeks of the year?  That’s where you are driving behavior.  This is where things like trust in the company, justice, and other things become important.  How many people get up every day and think: this is my rating?  How am I going to react today?  People want to feel their work made a difference and are energized when they get praises and positive feedback.  Why wait to do that only once a year?
​

Hopefully, this provides greater clarity on the issue.  See the seven-part list summarizing the guidance. The biggest challenge is not with performance ratings (although there’s room to make them more accurate), it’s the missing focus on the ongoing development and keeping employees motivated.  Dr. Smither also talked about the importance of trust and justice to get the full benefit of performance management.  There’s plenty of well-researched evidence-based guidance available on this topic.  We cover these and provide a comprehensive guide in our course on Performance Management.  Follow this link to learn more

Dr. Jim Smither is the Chair and Professor of Management and Leadership at La Salle University.  Previously, Jim was a Senior Manager/Group Leader in corporate human resources for AT&T where he was responsible for developing and validating employee selection programs for management-level positions.  He is currently a Consulting Editor at the Journal of Applied Psychology. He has edited two books (published by Jossey-Bass) that present research/evidence-based best practices in performance management.  He has consulted with over 55 firms in human resources and leadership development.