Keep the Performance Reviews!

Ok, I cover the market for solutions that help automate performance reviews.  So, if you think that makes me biased, so be it.  However, I thought it was important to post about a recent Wall Street Journal article called "Get Rid of the Performance Review!".  The author offers seven reasons why performance reviews are ill-advised and bogus.  I will offer commentary (and counterpoints) to each:

• Two People, Two Mind-Sets -- The author believes the boss and subordinate have different motivations.  The boss wants to discuss where performance can be improved while the subordinate is concerned about compensation and career advancement.  I think this true, but not irreconcilable.  That is the whole point of pay for performance.  An employee should be concerned about improving performance if they are concerned about compensation.  Likewise, a manager should be concerned about development (including career development) if they want subordinates to improve performance.
• Performance Doesn't Determine Pay -- The author believes market forces are the primary determinant of pay.  Certainly, market forces do drive total compensation as well as the ability to pay bonuses.  However, most organizations (studies show nearly 80% in the US) have some sort of variable pay program.  Companies that truly differentiate performance (that is a pre-requisite -- if there is grade inflation, all bets are off) can also differentiate that variable compensation.  This is also an important point.  Some companies that do differentiate performance do not always differentiate bonuses and stock allocation.  However, a well-conceived and administered pay-for-performance program can motivate employees to improve performance.  I do not buy the argument that a performance review is a "cover story" to justify pre-determined pay.  There are certainly boundaries placed on pay based on market forces, but in many organizations there is a lot of flexibility within those boundaries.
• Objectivity is Subjective -- For some jobs, performance evaluation is actually quite objective.  There are easily quantified for metrics to evaluate performance in jobs like call centers and sales.  The challenge is that for many jobs there is going to be a subjective evaluation.  The author is correct to point out that there are always going to easy graders and hard graders.  However, that is why it is important to at least try to make sure that managers are using the same criteria for evaluation.  It is also why it is so important that organizations do calibration (are you listening vendors?) to minimize grading differences.  The author I think misses the point on the 360 reviews however.  360 assessments are typically used for development, not performance appraisals.  Some organizations certainly get feedback from multiple people in determining performance, but that is different (and certainly that feedback has to be taken with a grain of salt as the author points out).
• One Size Does Not Fit All -- I do not know where to begin on this one.  Performance reviews have structure so that there is a consistency in terms of what is evaluated in each position.  However, what is actually put in the review and the subsequent conversation around can be individualized.  In addition, you do have to evaluate the right competencies and goals for the job and the individual.  I think the author and I agree on that point. I have often told clients (and others) that garbage in = garbage out.  If a reviewer enters boilerplate information into a review that is the same for everyone and does not take the employee conversation seriously, you will get poor results.  If the manager is not assessing the right competencies or the wrong goals are set, you are going to get sub-optimal results.  However, you would still get the same poor results even if you did not have performance reviews because the manager in question is a poor manager.  Performance reviews are not the problem.
• Personal Development is Impeded -- I am sorry but I just do not get the justification on this one.  As I understand the author, subordinates will not turn to a boss for development advice because they do not understand them well-enough (I guess because the subordinate does not think the boss has captured reality in the performance review).  In addition, the author asserts subordinates do not want bosses to know their weaknesses because it could come back to haunt them in a future performance review.  There is certainly plenty of research that shows that the relationship between subordinate and boss is the important factor in employee engagement (and ultimately productivity and performance).  There are plenty of bosses who do performance reviews and have good relationships with their employees.  They are not mutually exclusive.  If a subordinate does not trust the boss, regardless of whether or not there is a performance review, there are going to be issues.
• Disruption to Teamwork --  The author contends that the performance review is one sided and gives all of the power to the boss and creates a "conflict of interest" in effective teaming.  This certainly can happen.  However, it does not have to happen.  I know one professional services firm that allows any person at any level to provide performance feedback for any other person regardless of level.  It works in their culture.  360 assessments and employee climate surveys can also be useful in identifying managers that are too one-sided.  To look at the performance review alone is does not paint the full picture.
• Immorality of Justifying Corporate Improvement -- The author thinks it is immoral to maintain the facade that annual pay and performance reviews lead to corporate improvement (leading to more bogus than valid activities).  The author cites no authority for this assertion.  There is ample research including the annual Watson Wyatt Human Capital Index studies that show companies that utilize best practices in "Total Rewards and Accountability" have significantly higher total return to shareholders than those who do not.  Obviously, there are companies that "talk the talk", but do not "walk the walk".  There is plenty of other research that show employees do not feel performance reviews are fair, consistent, etc..  It does not mean that performance reviews are not worthwhile, it just means that you have to have a healthy performance-based culture to make it work.  If you try to do pay for performance and align it with corporate objectives without the right culture, it is probably doomed to failure.
• The Alternative -- Performance Previews -- The performance preview is future-oriented.  It focuses on what supervisors expect from subordinates and how they can work together to accomplish what both want to see happen.  It is not an annual process.  Rather, it takes place anytime boss and subordinate are not working well together.  I do not see this as a wholly new idea.  It groups some good practices together under a single banner (which is not a bad thing).  Many organizations do goal setting at the beginning of the year to level set expectations of supervisors with subordinates.  If there is training or other resources required to meet those objectives, those issues can be identified up front and addressed.  However, the author makes a good point.  Sometimes, this is the only time these issues are discussed.  They are not discussed again until the end of year performance appraisal (or maybe in a mid-year conversation).  However, the performance appraisal should not be the be all end all of the performance conversation between managers and employees.  Informal feedback and discussion are important and supplement the more formal performance appraisal process.  I do not think you have to throw out performance appraisals to get the benefit from informal performance discussions.

Even though I am speaking up for performance reviews, I do recognize that they can cause more problems than they solve.  I also realize that many organizations do not have the right culture, systems, structures, and trust (between line employees and management) to get good results from performance appraisals.  However, I do not think that means we need to throw out the performance reviews.  It just means there is room for improvement.  I do believe accountability is important.  I do believe compensation and rewards can drive specific actions and behaviors (you just have to be careful they are the ones intended).  So, it would be a mistake to get rid of performance reviews entirely.

What do you think?  Are performance reviews a good thing?  Do you think the idea of performance previews is a better approach?

Comments

[Gopal Shenoy]

Jim,

I could not agree more with your comments. As a manager and an employee, I have always made sure what Culbert calls "performance previews" (we call it weekly one-on-ones) happened every week. This way there were no surprises when it came to performance review time. So it is not that one has to be replaced with the other. I think both can happily co-exist.

Now where do I see the maximum benefits of performance reviews as an employee and a manager. It is a nice documentation on what has transpired throughtout the year. Just like how one goes back and looks at photographs of past years, I look at performance reviews as where have I been, what have I accomplished along the way, what new things have I learnt and what corrections where I asked to make over the years by my managers etc. Without such a record, all I would have is my resume which indicates some of my key accomplishments.

If managers don't know how to manage employees correctly, that is a different problem. Don't blame performance reviews for that. It is after all a tool. Any tool fitted to a lousy process is going to fail. Even if you completely get rid of performance reviews, you are still going to be left with managers not knowing how to manage their direct reports.

Gopal

[Steve Bonadio]

Good post Jim. I have to admit that I was scratching my head after reading the WSJ article. I wrote a rebuttal of sorts on the Softscape blog (http://blog.softscape.com/?p=32) but the bottom line is that I’m not clear that the author of the WSJ article knows much about best-practices in performance management. He makes zero mention of goals or competencies – how can you have an educated discussion about performance management without diving into goals and competencies? His views on pay-for-performance seem to ignore the empirical data on the subject. And, he has some rather faulty views on employee development (e.g., 360s are used for performance, the boss is "responsible" for employee development). I agree with you that there is room for improvement. But to call for an end to performance reviews without demonstrating a clear understanding of the issues at hand seemed a tad inflammatory to me. Or as I put in the Softscape blog, naïve populist sentiment.

[Jon Clemens]

Good post. I agree with your comments and with the comments of those that responded above. The problem is not performance reviews, but its with those companies that perform performance reviews as a backward-looking only, one-sided view of the employee. Instead of getting rid of performance reviews it would be much more beneficial for companies (and their employees and managers) to adopt best-practices in this area.

Having said that, it is true that for many companies that do not have best practices and who have not automated their performance review process that the value received through the reviews may not be worth the effort. A far better call to action, in my biased opinion http://www.jakoba.com/learn.htm , would be to adopt an automated tool that not only alleviates much of the administrative headache of reviews, but that also embeds best practices within the process. Adopting such a tool can completely change the value proposition of reviews around where the benefits are obvious and pain is much reduced.

[David Vanheukelom]

Jim,

Great post and I would second your commentary.

Most organization that I work with, and this include both management as well as employees, understand and appreciate the value a well defined performance assessment and development process can offer an organization in business alignment, productivity, leadership development, top talent retention, and so forth. This would include the concepts of meritocracy and pay-for-performance whereby again my experience is positive with respect to employee alignment and agreement on principle and strategy. One of the largest challenges I see with respect to the organizational difficulties inherent in creating a performance culture is on the acceptance of performance criteria that is objective and measurable by both managers and employees (SMART objectives are the usual standard that is discussed). As you noted above, a number of job roles can have very clear and measurable performance standards, but there are also a number of roles whereby concrete or 'calculated' measures of performance are not as easily discernable. This is a fundamental and prevalent perception that I would argue is a cause of a number of the arguments as stated in the WSJ article and many others - including the impact of two-minded bias, bias in 360 degree as noted by the author, subjectivity verus objectivity, and lack of accountability. In addition, most managers I've met are passionate about the important role they play in managing, developing, and assessing their talent objectively, but struggle to find objectivity in measuring and discussing performance with their employees; with the result quickly becoming distrust with the process, avoidance of responsibility and employee communication, and questionable tactics in evaluating employees. Moreover, organizations need to understand subjectivity in a manager-employee relationship absolutely exists, but this should not be a cause for abandonment of either process or trust. What does need to occur is considerable management of cultural change, business and organization strategy, and communication of trust between managers and employees - which would include mutually agreed upon peformance objectives and criteria for measurement between the manager and employee to which requires a culture and relationship of trust. Finally, a successful performance culture requires an organizational-wide commitment to cultural transformation, the management of social and human capital across the business, and a fundamental shift in leadership development for managers - and employees.

[Donna Ronayne]

Culbert is accurate in outlining many of the issues organizations still face when it comes to performance reviews. The idea of replacing reviews with performance previews points out how goal setting and establishing expectations should be included as part of the appraisal process. That said, the “preview” is only one step required to overcome the challenges of performance reviews. It’s not about getting rid of reviews altogether, but taking the action needed to create a talent management process that addresses each of these issues and creates true value for everyone involved. Lots of articles on best practices for talent management can be found here: http://www.halogensoftware.com/resources/reference-library/

[Working Girl]

I think we should keep the reviews but with less emphasis on the automation. I’ve heard it argued that canned feedback is better than nothing. This is probably true but there is sufficient research to indicate that personal recognition makes people feel valued, which can be more motivating than cash. I would therefore argue that better than nothing is not good enough.

Talent management is a human problem first and a software problem second. So far the leading talent management solutions have not solved the problem of poor management, which seems like a pretty important problem to solve if you want to attract, retain and motivate good people. Note that poor management can manifest in many ways, from failure to weed out poor performers to vague, confusing or conflicting corporate objectives, and automated performance reviews in a poor management environment will at best create some baby steps in the right direction and at worst bog everyone down in pointless activity.

Not that this is a criticism of talent management solutions because in all fairness, management of people isn’t a software problem. The criticism is that leading talent management solutions have implied that automating talent management processes is some sort of recipe for better management. The real solution is better management, which can’t be installed.

A performance review done well, with an emphasis on the people involved rather than complex cascading goals and what percentage of reviews are complete, can be tremendously value adding for everyone involved. So, I agree we should keep the reviews. But I think that the value add comes when people have an honest, constructive interaction that results in the right people being recognized and/or rewarded, not from the automation.

[Ken H]

Performance management systems suffer from chronic dissatisfaction in most organizations. This is due to real management and measurement challenges associated with 'performance'. Trying to simplify these challenges just ensures continuing dissatisfaction.

Inevitably, such systems call for rating people on their performance, and we use various types of scales in making those ratings, often mixed together. "Your performance is above average." Or, "You consistently exceed expectations." Or, "Your performance is strongest on factors that just aren't among the most important compensable factors." Or, "The strengths and weaknesses you demonstrated during the past year were: ..."

These factors describe quite different elements of work performance, and often don't blend together very well. Concepts mixed together:
• Meeting expectations
• Average performance
• Strengths & weaknesses
• Compensable factors
• Performance versus results

Who said these are identical factors? They measure very different things, and performance management systems fail due to muddy thinking and incompatible assumptions among these five factors.

Let's address them one at a time;

1. Meeting Expectations. Why should we think anything other than most people will meet or exceed expectations? Groups of people, including supervisors, generate powerful self-regulating norms around expectations, all for very good reasons. The norms include:

o Individuals who exceed expectations get more responsibility and assignments added to their role. It's called 'career advancement', and represents a rational allocation of tasks by supervisors to those most ready to perform them.

o Individuals who fall short of expectations get subtle, but very intense pressure to measure up. They stand out and the issue demands correction. This is where supervisors make many mistakes: they fail to confront those who don't measure up to expectations, either by failing to make expectations clear, or by failing to correct the situation. When this happens, others in the group will make compensating action: picking up the slack or giving up in frustration, thereby lowering morale and even raising turnover. In other words, having the staff measure up or more than measure up is also a core measure of supervisory effectiveness.

o Then, there's the internal striving each of us has to measure up. To use the psychologists' term, measuring up is important to our sense of self-efficacy. If we don't feel as though we're meeting expectations, we ask to clarify them, and seek to adjust the workload or our performance in order to better measure up.

o Supervisors who know how to do their job well design and structure work so that people in the various roles can do the job successfully. It's rational, appropriate managerial behavior. If managers do their job reasonably well, why would we expect people to fail to meet expectations? Rather, we would expect a large majority to do the job well. This phenomenon helps explain why jobs are continually in flux and 'standard' jobs never stay the same across an organization: managers routinely adjust the jobs so that the people in them can perform them reasonably well (meet expectations). Every manager does this according to the talent available, not according to some rigid structure of job descriptions.

o Therefore, if the scale of performance ratings uses 'Expectations' as its mileposts, the resulting distribution of ratings should look like an HR Executive survey from a couple years ago, with most people "Meeting Expectations", a small number "Do Not Meet Expectations" and a fair number "Exceeding Expectations" as they strive to advance. A better question is whether "Expectations" is the right scale to use, and for which purpose. It's probably exactly the right scale to support individual motivation and higher levels of effort, but is it the right scale for distributing tangible rewards (pay increases, bonuses, etc.)?

2. Average Performance. What does this even mean? Technically, it's the mean, median, or modal statistic on some meaningful measure. We all know that. But, what measure? I suspect people generally think about the distribution of results generated by individuals in a group when they think about 'average'. But, they might mean capability levels or talent. Or, they could be thinking about future potential when they think about 'average'. In managing performance, 'average' is perhaps one of the most dangerous, even wrong-headed notions to apply. A corollary to the points on expectations is that managers are charged with achieving high performance if at all possible. What does an 'average' performer look like in a high performing organization? And, what message does that 'average' employee take away from the discussion? Just as irrelevant, what does an 'average' performer look like in a poorly performing organization? And, is it a positive message to be 'above average' in a failing organization? Maybe, but there certainly are limits to the value of any interpretation of a comparison to average in either circumstance.

3. Strengths and weaknesses. This is a very important developmental concept ("You need to improve your listening skills"), but it too, is more complex than apparent on first blush, and what place do these have in performance management anyway? If the performance management process includes an emphasis on learning and growth, then strengths and weaknesses are perhaps the most central notion to emphasize. But, strengths and weaknesses are only indirectly related to meeting expectations or to achieving results. More importantly, despite the overwhelming urge to focus on weaknesses (fixing what's broke), powerful reasons mitigate toward a focus on strengths.

o People naturally gravitate toward strengths - we like to be successful and feel on top of our work; therefore, we try to organize our work to do more of those things at which we're proficient.

o Higher levels of motivation emerge when people do more of what they're good at. They receive more positive feedback as a natural outcome. Higher levels of motivation correlate strongly with higher levels of productivity.

o People are extremely adept and creative in adopting work methods that leverage their strengths, and compensating for or working around their weaknesses.

o Supervisors assign and direct tasks based on the differential strengths on their team. After all, who would assign a critical task to someone who is incompetent to fulfill it, or even to someone relatively less competent than another team member? Of course, that's exactly what clever managers do in order to stimulate growth and development - take a risk on someone by giving them a stretch assignment. But these types of risks are normally carefully calculated so as not to jeopardize organizational performance.

o Ergo, developing weaknesses should be a secondary developmental focus, and aimed at those skills that raise barriers to very specific work requirements. I.e., our poor listener really needs improvement if that poor listening gets in the way of effective job performance. A focus on weaknesses may be essential to addressing performance barriers and also to enable career development. But if it doesn't get in the way too much, let it be, at least in the short run; leveraging strengths is just plain far more powerful.

4. Compensable Factors. Ah, here's the area where specific ranking, both absolute and relative (how do people compare?) is very important. Managers need mechanisms for determining who gets what level of salary increase and how much differentiation to implement among people in their salary increase and bonuses. But, herein lays the tension between results and performance.

o The psychologists stake out a clear position on individual performance: it's what people do, not what they produce. Think of an actor in a play. The performance is what the troupe does each night on stage, and the actor's performance is her behavior among the troupe in playing her part. Period. The results (was it an enjoyable play, did audiences fill the theater, did critics praise or pan the play) speak to multiple factors affecting the outcomes: the play itself (the text), the direction, the set & other production factors, the performance of the actors, and even such completely independent variables such as the local economic condition (affecting how many people buy tickets), the marketing & promotion efforts, etc. The performance of our actor is certainly vitally important, but is only one factor among many affecting the results of the play. How many movies or plays have we seen in which a leading player acts brilliantly in an otherwise failed production?

o What, then, is 'compensable'? The factors having the greatest effect on compensation are the structural ones built into the job description: scale, scope, complexity, external talent market pressures, etc. These factors derive from the design of the job and its value in the external labor market, and not from the people filling it. The performance (either behavioral and/or results) of individuals do affect base compensation but much more 'at the margin'. I.e., a high performer will earn an annual increase at the high end of a range of merit increase, say perhaps 5% to 7%; a poor performer or typical performer will earn less, anywhere between 0% and 4%. Note the difference: 1% - 7% maximum differentiation in any year, outside promotional increases or market equity adjustments.

o Identification of compensable factors leads naturally, then, to a focus on promotion, advancement, and job enrichment/expansion - those factors driven by structural changes in the job. As such, compensable factors should lead to an emphasis on the distribution of opportunities, rather than on current performance. While the distribution of opportunities is affected by current performance and performance track record, long term potential also plays an integral role in determining who gets what opportunities and how quickly.

5. Performance vs. Results. The most prevalent contemporary trend in performance management calls for identifying both types of measures - behavioral measures (individual performance) and results measures (the outcomes of performance) - in establishing performance targets and performance management system parameters. Important qualifiers need further commentary regarding these two perspectives.

o Behavioral measures of individual performance are most frequently based on competency models, and best practices include behavioral anchors for different levels of performance on the competencies. However, generic competency definitions often leave these important measures less meaningful for any individual. If these are to serve well as a basis for measuring individual behavioral performance, then they must be adjusted to be particular to the role and person in question.

o Another variable affecting the value of measuring behavioral performance is the means of measurement. 360-degree evaluations can often overcome errors that managers might make, most importantly two: first, that many important behavioral dimensions are demonstrated in interactions with peers and direct reports, and not to the boss; second, the boss may have a biased perspective on the behavioral performance of an individual, including either positive or negative 'halo' effect, past performance bias ('this individual has been a strong performer for several years, therefore they must still be'), etc. But, while these types of measures are very meaningful for identifying strengths and weaknesses (within person relative measures), they don't reliably or accurately pinpoint differences among various individuals. Consequently, 360-degree measurements are great for supporting a developmental purpose, but misleading and possibly damaging, for differentiating performance and rewards among individuals. (Among other flaws, e.g., the research shows a systematic inflation of 360-degree ratings when rewards and decisions are at stake, as compared with ratings used solely for development.)

o Yet another challenge with behavioral performance is that the individuals rated view any evaluation of behavioral performance as too subjective and therefore subject to favoritism and bias, no matter how well-designed the system or well-intentioned the managers. People often like 'hard' measures of performance, both for objective-setting and performance evaluations. Hard measures remove the equity, justice, and bias problems too often associated with measuring behavioral performance. Or do they? Hard measures of performance are rarely under the sole control of the individual, and often are heavily affected by the business processes or even the strategies of the organization. Further, completely external factors often heavily affect performance on hard measures. E.g., does anyone perform well in automotive sales in today's environment?

The bottom line: be clear about what objectives you're trying to meet with the performance management process, and be sure to align the measurement process(es) to fit those objectives. Over-simplification of performance management will inevitably lead to either unintended negative consequences or at best, chronic dissatisfaction with the process.