Grote Consulting Resources

Books to Help With Performance Appraisals

Passing Judgment

PDF Version

By Dick Grote

A universal complaint in organizations involves the ineffectiveness of their performance management process. Performance appraisal grumbles arise from all quarters. Employees often see themselves as victims of a subjective, failure-oriented procedure that focuses only on highlighting faults and flaws. Complaints come from managers who feel awkward as assessors and are uncomfortable having hard conversations. Human resource professionals who are charged with making sure that these systems are designed correctly and executed effectively get blamed when things are harder than they ought to be and don’t go as they should.

And you, the corporate executive, aren’t happy either. You need accurate and solid data on the quality of talent and performance of members of the organization, but you’re suspicious that your performance management system just doesn’t deliver the goods. As important as performance management is, no one seems to like the current state of affairs.

And performance management is important. It serves a vital — an irreplaceable — function. Effective performance management systems allow organizations to make critically important decisions correctly. Decisions like:

  • How should rewards should be allocated? Who should get a big raise, who should get nothing at all?
  • When a vacancy arises, who should be tapped for promotion? Do we have excellent candidates already ready and waiting, or do we have to go outside to get the talent we need?
  • What’s the depth of our talent pool? Do we have the people we need to meet the demands of the future?
  • What is the relative strength of talent across our organization? Are there pockets of excellence and pockets of mediocrity within the company?
  • Who are our best performers, the people who are outstanding performers in their present positions with the potential to take on more demanding roles? Do we have retention strategies in place to make sure that they don’t leave us?
  • Who are our worst performers and what do we need to do about them? Which ones can we salvage? Which ones should we cut loose?

If the performance management system isn’t intelligently designed and skillfully used, the answers to these critical questions will of course be wrong. Worse, the questions won’t even be asked, and the organization will stumble towards achieving its strategic goals without knowing whether it has the talent it needs to assure their achievement.

For years the argument to simply abolish performance appraisal altogether has been made, citing the apparent dubious accuracy of the data generated by appraisal procedures, the ill-designed forms, and the discomfort of managers and employees alike in using the process. But redesigning forms, training managers in how to be better communicators and restricting appraisal data only to that which can be empirically proved simply won’t work.

A few companies have tried eliminating their formal performance appraisal mechanisms, but have discovered that things then only seem to get worse. Goal-setting doesn’t happen. Managers hold even fewer feedback discussions without the goad of a formal requirement. Employees don’t know where they stand. Salary administration gets more cumbersome, and companies discover that they’ve ended up with a bunch of itty-bitty unconnected systems that — joined together — look just like the conventional appraisal process they thought they got rid of. A few years ago a Merck task force explored all possible alternatives to their traditional performance appraisal system. They finally decided to stick with things just as they were.

The reason that simply abolishing performance appraisal won’t work is that the focus of all the complaints — bad forms, low-quality data and inept discussions — is not the problem. The problem is that everyone involved with the performance management process — executives, assessors, recipients, and human resource management professionals — don’t understand some of the most fundamental concepts underlying performance management. Worse, they believe a set of myths that actually prevent their performance management procedures from operating successfully no matter well the managers have been trained and the forms designed. Only when these myths are brought to light and eradicated can performance management systems deliver what they are capable of and give organizations what they desperately need: valid, workable information that allows good business decisions to be made.

What is a performance appraisal?

One of the most basic causes of performance appraisal failure is that so few people understand just what a performance appraisal is. Listen up: A performance appraisal is a formal record of a manager’s opinion of the quality of an employee’s work.

The operant word here is opinion. Performance appraisal requires a manager to render his opinion about exactly how well an individual performed. It is not a document that can be empirically tested and proven. It is not the end product of a negotiation between the manager and the individual. It is a record of the manager’s judgment about exactly how good a job Joe or Jane has done over the past twelve months.

As long as that judgment is rendered based on the manager’s honest assessment of how well the job has been done without personal prejudices and biases coloring that assessment, the performance appraisal is virtually immune from a disgruntled employee’s legal challenge. The notion that delivering an honest but negative performance appraisal will lead to an appearance on the witness stand in a discrimination lawsuit is an unfounded myth. Courts have consistently upheld the right of organizations to make business and employment-related decisions based on managers’ opinions as recorded in performance appraisals even when those opinions are based only on the manager’s observations, judgment and experience without provable, quantifiable data to support those opinions. This deference by the courts to employers has been stated as the business judgment rule and has been acknowledged by the courts plainly: “We do not assume the role of a ‘super-personnel department,’ assessing the merits or even the rationality of employers’ nondiscriminatory business decisions.” The appraisal is a record of a manager’s opinion. If the employee and the manager disagree about that opinion, the manager wins.

Understanding what a performance appraisal is — a formal record of a manager’s opinion — eliminates another source of mischief that causes performance appraisal discussions to go awry. In the management seminars I conduct on making performance management work, I hear a frequent question, usually stated as, “I’m about to discuss a performance appraisal with one of my subordinates. The appraisal is honest, but there are several things I’ve written that I know she’ll disagree with. How do I go about gaining her agreement?”

Stop! Another myth of performance management is that the objective of the performance appraisal discussion is to gain the employee’s agreement. It’s not. If the manager has applied tough-minded, demanding standards, it’s unlikely that the individual will agree. That’s OK — the objective of the meeting is not to gain agreement. The tougher the manager’s standards, the less likely it is that agreement will occur. The objective of the performance appraisal discussion to get the individual to understand the reason his performance was rated the way it was, not to gain his agreement on the accuracy of that rating.

The infinite human capacity for self-delusion

While the individual’s personal insights into the quality of his own performance may be a useful data point for the manager in creating the formal performance appraisal, another common myth holds that asking the employee to complete a self-assessment using the company’s form, or including the perspectives of others gained through a 360-degree feedback system, is a good idea. It’s not. It’s a bad idea and needs to be stomped out.

In their well-known article, “Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments,” Cornell University researchers Justin Kruger and David Dunning report that those who are incompetent performers are also incapable of assessing the difference between good and bad performance. As they put it, “When people are incompetent in the strategies they adopt to achieve success and satisfaction, they suffer a dual burden: Not only do they reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the ability to realize it. Instead, they are left with the mistaken impression that they are doing just fine.”

As one senior executive describing his company’s experience using a forced-ranking procedure to identify its A, B, and C performers put it, “The As are afraid they’ll be considered Bs, the Bs are scared they’ll be seen as Cs, and all the Cs are convinced that they’re A-players.”

Research consistently demonstrates that individuals are notoriously inaccurate in assessing their own performance, and the poorer the performer, the higher (and more inaccurate) the self-assessment. Research by the firm Lominger, Inc. indicates that “The overall correlation between self-ratings and performance was .00. The most accurate rater by far is the immediate boss.”

In July 2007 BusinessWeek surveyed 2000 Americans in middle management positions and above, asking them the question, “Are you one of the top 10% of performers in your company?” Not one of the subgroups in the survey had less than 80 percent of the respondents answering the question affirmatively. 84 percent of all middle managers reported that they were in the top ten percent of performers in their company. Executives — the most deluded cluster by far — provided a Yes response in 97 percent of those surveyed. Lake Wobegon, indeed!

“Know thyself” may be good philosophical advice, but in assessing how good a job you’ve done, your boss knows better than you do.

Bosses can’t run unchecked, though. While the judgment of the individual’s immediate supervisor is the foundation of the appraisal, wise managers discuss with each of them the way they will measure team members’ performance as part of their start-of-year goal-setting discussions. If subordinates are geographically removed, good managers ask them to take responsibility for making sure that the manager has the data he needs for accurate evaluation at the end of the year. If the employee’s job involves customer service, they jointly develop a plan to survey the employee’s customers about their satisfaction with the service provided. And requiring appraisers to get their own boss to sign off on every appraisal written is becoming a routine (and wise) practice.

Asking employees to write self-assessments using the company’s appraisal form generates another problem. Earlier we defined a performance appraisal as a formal record of a manager’s opinion of the quality of an individual’s work. If the individual is asked to complete an appraisal form assessing his own performance, not only is he likely to be woefully inaccurate, but simply the act of filling out the form sets up the understandable but erroneous expectation that he and his boss will then compare and contrast their two assessments, the purpose of the appraisal discussion being to come to common ground. No, it’s not. The purpose of the meeting is for the boss to make clear what his evaluation is and provide the rationale behind it.

Senior executives need to be full participants in the performance management process themselves, both on the giving and the receiving side. Quantitative P&L results tell a lot of the story of how well the executive is performing, but top dogs need the benefit of performance management just as much as people who actually work for a living. Peter Drucker’s thirty-year-old concept of creating a “manager’s letter” is probably still the best performance management technique to use with senior executives. Annually each executive writes a letter to his superior spelling out the objectives of his own job and those of the superior’s job as he sees them. He then sets down the performance standards he believes are being applied to him. He lists the goals he intends to accomplish, the major obstacles to achieving them, and the support he needs from his superior and the company. Finally he spells out what he intends to do during the next year to reach his goals. If his superior accepts this statement, the “manager’s letter” becomes the charter under which the executive then operates.

Differentiation — And Why We Don’t Get It

Whenever I conduct an Executive Overview of best practices and new directions in performance management, I hear corporate executives consistently voicing two concerns. One involves building a culture of accountability, an environment where their subordinate managers honestly assess their people against tough and demanding standards, tell the truth, and take personal responsibility for their assessment judgments.

Their other concern involves differentiation — getting managers to make real distinctions in the performance of their subordinates and go on record about who their stars, their Steady-Eddies, and their laggards are.

Assuring accurate differentiation is vital for talent management efforts to succeed. But attempting to increase differentiation by providing managers with more rating levels — a common approach — is counterproductive.

What’s the primary cause of lack of differentiation in performance appraisal results? I find that it is allowing the use of decimal points for final ratings. Assuming that 1 is Unsatisfactory and 5 is Outstanding, if the final performance appraisal rating allows the use of a decimal point, then the final rating of almost everyone in the organization will be between 3.4 and 3.6. Since the 3 rating equates in most people’s minds to “average,” allowing a decimal point final rating permits spineless managers to tell everyone that they’re better than average — some a bit more so, others a bit less. Allowing the use of decimal points turns a 5-level rating scale into a 50-level scale (1.0, 1.1, 1.2 . . . 4.9, 5.0). But allowing more rating levels will produce less real differentiation.

The problem is compounded when companies provide even more rating levels than the usual five. For example, Shell of Canada has a rating scheme of 0 to 1.5 — in effect, a 15 level rating scheme in that there are 15 possible scores ranging from 0.0 to 1.5. What’s the outcome? Performance appraisal ratings cluster around 0.8 and 0.9.

Human beings simply don’t have the capability to distinguish among 15 levels of performance, let alone 50. Five levels of performance are about all that any of us can distinguish. We can probably distinguish among performance that’s exactly what we’re looking for, performance that’s significantly better or worse than what we expect, and then performance that is either breathtakingly excellent or appallingly dreadful. There’s your 5-level rating scale. If executives demand that managers use it as designed and eliminate the decimal point cop-out, it will provide all the differentiation needed to make valid and wise performance-based decisions about compensation, promotion, who should be terminated when a RIF is necessary, who special retention efforts should be directed toward, etc.

In fact, TQM devotees argue that only three levels of performance can actually be empirically determined: (1) that which is in spec, or “Meets Expectations,” (2) that which is out of spec on the positive side, or “Exceeds Expectations,” and (3) that which is out of spec on the negative side, or “Fails to Meet Expectations.”

But a more important concern than permitting the use of decimal points or allowing more than a five-level scale is that giving managers a plethora of choices encourages them to turn performance management into an arithmetic problem. Most managers would rather endure a root canal than to have to evaluate performance (and make performance-related decisions and hold performance discussions) using a small number of clearly-labeled categories.

With a decimal-point final rating scale, it’s easy for managers to duck hard decisions by treating performance management as a numbers game. And they’ll do it both in terms of assigning a rating for compensation purposes (“Well, it doesn’t really matter whether I give Betty a 3.4 or a 3.6 . . . ”) and they’ll do it again when they talk with the individual about the rating and accompanying raise (“Gee, Betty, I wanted to rate you higher but only a few people can get the top rating / my boss wouldn’t let me / HR wouldn’t let me / the system wouldn’t let me . . .”) Performance management is not an arithmetic problem. Allowing managers too many rating levels will lead them to cluster everyone at the point that’s just above average.

The Myth of Quantifiability

Another false belief held by people throughout organizations that in order for a performance appraisal to be objective, managers are required to find numerical, countable units to back up their assessments. That’s nonsense.

“Objectivity” has nothing to do with quantifiability. Objectivity has to do with being free of personal prejudice, being factual, basing appraisals on observable phenomena, like the way the person goes about performing his job.

Consider — how do you measure the performance of a pianist? The number of notes she played? How do you evaluate the performance of a painter? The number of pictures he paints or sells? Do you evaluate the performance of a priest by the number of confessions heard or the number of souls who gain entry to heaven (a valid measure perhaps, but accessing the data is problematical). A few years ago I was engaged by the National Security Agency to help NSA develop a new performance management system. NSA is probably the most clandestine of all of America’s intelligence organizations, keeping our country safe by intercepting and decoding the messages of our enemies. The National Security Agency hires more linguists and translators than any other organization on earth. How do you evaluate the performance of a translator?

The obvious, easy, and wrong answer is: Number of documents translated. But while that’s easily quantifiable, it misses what’s genuinely important. What’s important is the ability to capture nuance.

The ability to capture nuance isn’t hard to evaluate accurately. Just take an intercepted document written in some foreign language and give it to two translators. Then take their two translations to a native speaker and ask, Which one got it right? The native speaker will read the two documents and then comfortably say, “This one translates each word accurately. But this one — this one captures what the writer really intended.” That evaluation is certainly objective — but there’s nothing quantifiable about it.

What does it mean to be objective? It means to be uninfluenced by emotions or personal prejudices. It means to be fair. It means being factual, basing our assessments on things we can actually see. A person’s job performance is certainly observable. Managers can watch how someone does his job and assess how well he’s done it.

If a manager has access to numerical measures of the quantity of work the person did or has some quantitative index of quality to support her evaluation — terrific! Use them. But as long as the boss can provide solid examples to back up her assessments and ratings then the appraisal is objective, even if she can’t come up with countable units. Even if some people can’t be shaken loose of the foolish belief that if you don’t have numbers to support your assessment then it is necessarily “subjective” — they’re wrong, but we can tolerate that error — realize that people desperately want this so-called “subjective” information. People want to know their supervisor’s opinion of them and how they’re doing. They want honest answers to their most important questions: How am I doing? Are you pleased with my work? Do I have a bright future here? There is no number that can answer those questions.

Fewer jobs today produce the kind of quantifiable results that lead to easy evaluation. The outputs of knowledge workers are harder to assess than those of widget-makers. Even in P&L positions where results are clear and performance seems easily quantifiable, just making your numbers is no longer enough. Companies demand not only results, but generating those results in a way that reflects organizational values. GE and PepsiCo, two companies famous for decades of attention to performance management, are good examples of this dual focus on both results and behavior. When he was CEO of PepsiCo, Wayne Calloway said, “It’s easy to get fired from PepsiCo. You’ll get fired if you don’t make your numbers. You’ll get fired if you lie. And the fastest way to get fired is to lie about your numbers.” In its Session C assessments these days, GE has moved away from the harsh “Fire the bottom 10 percent!” approach of Jack Welch. But GE is clear about what’s important, as demonstrated by its statement, “Values must accompany demonstrated operational excellence in order to advance.” In GE’s Session C’s today, managers are assessed along two dimensions: Performance and Values. Low performance but a strong demonstration of GE’s values produces an assessment of “Restart: Second Opportunity to Deliver Results.” But if you deliver high performance with low adherence to values? “Remove: Removals Reinforce Importance of Values Dimension.”

The actual results an individual produces may be quantifiable. But how she got those results, and the extent to which she modeled the organization’s values in generating them, isn’t subject to numerical measure. But behavior and adherence to values can certainly be described, and those descriptions of performance, supported by examples, are certainly objective.

Albert Einstein is credited with the insight that best makes the point: “Not everything that counts can be counted. And not everything that can be counted counts.”

Why do we do performance appraisal?

A common myth surrounding performance management involves its purpose. Ask almost anyone in an organization why the company has a performance appraisal system and their answer is likely to have something to do with compensation. It’s true, of course, that if we believe in pay for performance — a universally-accepted dogma in a capitalist economy — then there must be some mechanism to evaluate that performance so that rewards can be fairly meted out. But if organization members and leaders see the primary function of performance management system as serving as the handmaiden of the compensation system, then performance appraisal results will be skewed to serve the manager’s desire to do the best he can for his troops.

Many organizations build disincentives to honest assessment directly into the system itself. For example, if the performance management system tightly links the amount of the individual’s salary increase with her rating on the performance appraisal form, managers will be tugged between honesty and generosity. At a time of record gasoline and grocery prices, managers understandably want to be generous with their team. To do so they may decide to artificially inflate a rating in order to bump up a merit increase.

Of course, there must be a linkage between the individual’s performance and the individual’s rewards. The alternative is to use the step-increase, time-in-grade scheme used by the federal government, rewarding only longevity and the ability to keep from being fired, an approach that private industry rightfully scorns. But a quid-pro-quo, if-this-then-that approach to compensation where the performance appraisal rating determines to the tenth of a decimal point the amount of the increase the individual gets will corrupt both the performance management system and the compensation system.

The better alternative is to separate the compensation and the performance appraisal discussions and make the performance appraisal rating a major factor, but not the sole determinant, of compensation changes. While making this separation may add a bit of administrative burden, it will allow organizations to take into consideration such vital factors as the individual’s possession of a unique skill, the desire to decrease the likelihood of his quitting, a reward for being identified as a high potential individual, and other issues in making the final pay change determination.

“Additional administrative burden” may be the greatest source of resistance to making significant improvements in an existing performance management system, or to scrapping old procedures that have been around for years and getting a fresh start. “It takes so damn long!” managers grumble. Yes it does, if it’s done right, but it’s worth doing. It’s worth doing in the same way that budgeting, another time-consuming and burdensome process, is worth doing.

The real cause of complaints about excessive time requirements is often not that the actual procedures are unduly time-consuming, but that managers aren’t clear on exactly what’s expected, by whom, and when. Publishing a calendar of expected events in the annual performance management cycle can, just by itself, eliminate a lot of the grousing about “too much time.”

While the performance management system provides vital information for use in making compensation decisions, and promotion decisions, and development decisions, and termination decisions, the reason we do performance appraisal is ultimately because it is an ethical obligation of leadership. Consider this: every person who works for an organization wants the answer to two questions. First, What is it that you expect of me? Second, How am I doing at meeting your expectations? While informing decisions in other areas is a valuable by-product of the performance management system, the ultimate reason we do it is that we owe it to everyone who works for us to let them know what we expect and how they’re doing.

Courage — The Key Ingredient

The problem with performance management as it is practiced in organizations today is not one of bad forms and unskilled managers. A good manager with even the worst form will do a good job; a mediocre manager with even the most sophisticated system and the finest of training will still be mediocre.

That’s not to encourage improvements in training and system design. But the payoff will come in other places. It will come from assuring that everyone involved with the system understands why the company has it and what top management’s expectations are. It will involve confronting the myths that get in the way of performance management effectiveness.

Most of all, performance management success is at bedrock a matter of courage: the courage to tell the truth, hold to high standards, and be willing to say bluntly that Jane is better than Jim but not as good as Harry.

Dick Grote is a management consultant in Dallas, Texas, who specializes in helping organizations design effective performance management systems and build leadership excellence. He is the author of the management classic, Discipline Without Punishment, The Complete Guide to Performance Appraisal, and The Performance Appraisal Question and Answer Book. His most recent book, Forced Ranking: Making Performance Management Work, was published by the Harvard Business School Press.