Since 2007, my work has involved helping companies to maximize the value of a technology system that supports strategic human resource processes, including performance management (PM). This technology is largely agnostic when it comes to PM process design. It can be configured for PM processes that make extensive ratings or avoid ratings entirely, encourage elaborate goal cascading or completely forego the use of goals, focus on once a year PM events or encourage ongoing employee–manager check-ins, and so forth. In sum, the system can be configured in any number of ways.
The conversations I have with companies focus on determining what type of PM process will be the most effective in their particular organization. This will ensure the technology is appropriately configured to meet their unique needs. As a result of my role, I have discussed PM concepts with hundreds of companies crossing virtually every industry and located in almost every region of the world. Most of the companies I work with have thousands of employees, although I have worked with organizations with as few as 50 employees.
Given what I do for a living, I was quite interested to read the article by Pulakos, Mueller Hanson, Arad, and Moye (2015). My reaction to this article is mixed. On one hand, I applaud their efforts to explore new methods to increase the value of PM. On the other hand, I felt the article has failed to clearly define and systematically address the real challenges that impact PM effectiveness. It also criticized many common PM methods without acknowledging the value these methods provide for a great many organizations. The result is a highly skewed interpretation of what PM is, why it is difficult, and how to fix it. Nevertheless, the article does generate attention in an area that is in dire need of more research. So even if I do not always agree with how Pulakos et al. have discussed the topic of PM process improvement, I am glad they have generated the discussion
My overriding concern with Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) is a tendency to make sweeping statements suggesting that certain PM methods are universally good or bad. To their credit, they have stated that “there are no specific PM components that should always be retained or eliminated . . . they will depend on the way that PM system information is currently used; its value; and the specific strategy, PM goals, and needs of the given organization” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 30). But this statement appears after dozens of unqualified or weakly constrained comments about the effectiveness of different PM methods, such as “cascading goals are a significant time sink” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 10), “rating processes have grown overly complex” (p. 11), or “ratings add little to decision making” (p. 16). Many statements in the article suggest that the “best” way to do PM is to focus on coaching, whereas others suggest that use of certain PM methods related to goal setting and ratings are almost always doomed.
Observations made working with hundreds of companies to improve PM lead me to believe that there is no single way to “fix” PM because methods that work very well in one company may not work in another. For example, I have seen rigorous goal cascading methods that work very effectively in established retail organizations but that would probably fail miserably in high-tech start-up organizations. The best approach depends on the business needs and culture of the company, the nature of the employees and the jobs they are performing, and the skills and incentives given to managers. Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have acknowledged this, but they have not defined the criteria that influence when different PM methods are likely to succeed. Instead, they have presented several interesting PM methods that probably work very well in some companies but that I suspect would fail in others. For example, they have suggested using PM methods that emphasize giving employees a sense of autonomy, mastery, and purpose while downplaying ties to tangible decisions about employee pay and promotions. Disconnecting PM processes from pay decisions might work in some companies, but it could have a devastating impact in companies where employees are concerned about procedural justice regarding decisions that impact their financial stability (Colquitt, Reference Colquitt2001). In sum, although this article is full of “good ideas” for improving PM, what is really needed is more clarity about and research into the specific conditions in which these different good ideas actually work in practice.
There are many places in the Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) article where I found myself thinking, “yes, that is true in some situations, but it is certainly not true for all companies or jobs.” The following are some of the major areas about which I believe Pulakos et al. are only “half right.”
Performance Management Means Different Things to Different Companies
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have noted that “the promise of PM can only be realized if there is a shared mindset about what effective PM is” (p. 17). Yet Pulakos et al. have never actually defined what the term “performance management” means. For the purpose of this commentary, I define PM as
standardized and defined processes used to communicate job expectations to employees, evaluate employees against those expectations, and utilize these evaluations to guide talent management decisions related to compensation, staffing and development. (Hunt, Reference Hunt2014, p. 151)
Even this definition of PM can be interpreted in a lot of ways. I have worked with companies that think of PM narrowly, as a method primarily used to guide compensation decisions, and I have worked with companies that think of it broadly, as a process to develop and engage high performing employees.
Listed below are four common business objectives often associated with PM:
• complying with legal or contractual requirements that require clearly defining and communicating employee performance expectations and consistently evaluating and documenting employee performance on the basis of these expectations;
• engaging high performers by identifying employees who are making the strongest contributions to the organization and providing them with significantly greater rewards and recognition in the form of pay, career opportunities, and development resources;
• increasing employee performance and retention through enabling greater levels of coaching, constructive feedback, and dialogue around career development; and
• establishing consistent, well-defined processes and job-relevant criteria to guide pay, promotion, and termination actions.
All these objectives are valid reasons for implementing a PM process. However, a PM process that supports one of these objectives might not effectively support another. For example, the PM methods that Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have proposed would probably effectively support only one of these four objectives. So the suggestion that the methods that Pulakos et al. have outlined will fix PM is clearly limited to a fairly narrow definition of what PM involves.
PM Processes That Fail in One Company Can Be Highly Successful in Another
Much of the evidence Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have provided for the validity of their PM methods is drawn from a single company with a unique culture and set of business requirements. The article is largely a case study describing how this one company is seeking to change its PM process. Such case studies are useful because one can learn from what others are trying. But it is premature to state whether the PM methods advocated by Pulakos et al. will generalize to other companies.
Nor should we accept that Pulakos et al.'s (2015) criticism of different PM techniques will generalize to all other companies. To the contrary, many companies gain significant value from PM methods that Pulakos et al. have described with extremely negative terms such as “time sinks,” “detracting from performance,” or “overly complex.” For example, the evidence in Table 1 provides a marked contrast to many of the negative comments Pulakos et al. have made regarding the value of goal cascading, performance ratings, and annual goal setting. Because Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have justified the value of their PM process design recommendations using qualitative comments from managers (p. 28), I have used the same method here. I am not at liberty to share the names of the companies from which I have drawn these comments, but they are all direct quotes from line managers and employees using these PM methods.
Table 1 Manager Reactions to Performance Management Methods
The point is not to argue about whether one PM method is better or worse than another is. It is merely to note that many assertions that Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have made about the value of certain PM methods is not reflective of experiences found in other companies. Furthermore, the use of more evaluative PM methods such as ratings and pay for performance, which Pulakos et al. seem to have suggested are not worthwhile, has been shown to positively correlate with company profit and growth (Bloom & Van Reenen, Reference Bloom and Van Reenen2007). These more evaluative PM methods may be harder to implement, and managers and employees may not always like these methods, but many things that people find difficult and do not enjoy are often good for them.
Pay for Performance Is Very Effective if It Is Done Correctly in the Right Setting
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have cited the work of Daniel Pink in claiming that “carrot-and-stick rewards are not effective for driving performance or behavior in today's complex work environment” (p. 18). Aside from the questionable generalizability of Pink's theories of motivation across different jobs and employee populations (Hebert, Reference Hebert2010), this claim ignores research that shows pay for performance techniques to be very effective if applied appropriately in many settings (Rynes, Gerhart, & Parks, Reference Rynes, Gerhart and Parks2005). In sum, pay for performance, like most PM methods, may not work everywhere, but it can work very well in certain settings when used the right way.
There Is Nothing Inherently Wrong With Annually Based PM Methods
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have made several statements suggesting that annual PM processes are somehow inferior to methods that do not follow an annual cycle. For example, Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have stated the following: “the yearly review that provides backward-looking feedback is not only ineffective but also potentially damaging” (p. 16), “rather than set goals once per year . . . it is more effective to set ongoing expectations as needed” (p. 20), and “setting SMART objectives annually . . . ignores the dynamic and ever-changing nature of work” (p. 20). I agree with the elements of these statements that call out the value of forward-focused performance discussions and the need to update goals throughout the year to reflect changing business conditions. However, I disagree with the implication that using annual methods is an intrinsic part of the problem. For example, why should one expect that annual PM reviews are inherently backward looking and that ongoing PM feedback conversations are more likely to be forward looking? I suspect managers providing feedback in real time are very likely to be backward looking, calling attention to employee's recent mistakes, rather than stopping and planning out a forward-looking conversation. Similarly, because many corporate goals are set annually, it makes logical sense to establish initial employee goals annually with the same calendar as long as companies recognize that these employee goals are likely to be modified to reflect tactical and strategic changes that occur throughout the year.
Any PM activity can be effective or ineffective regardless of whether it is an annual review or a weekly check-in. The effectiveness of things like goal setting and PM reviews is a function of how well managers and employees carry these tasks out. It is not a function of whether companies choose to link these tasks to an annual calendar.
Pointing Out What Companies “Should Do” Will Not Fix the Core Problems With PM
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have emphasized that much of the value of PM lies in PM's ability to create better conversations about performance between employees and managers and that these conversations tend to happen outside of PM forms and process steps. I agree that this point is true, but it is not a particularly new observation, nor is it very helpful to companies that are seeking to create more ongoing performance discussions. Every PM training program I have seen emphasizes the importance of not treating PM as a once a year process. PM training typically includes statements like “there should be no surprises during a PM review” and encourages managers to engage employees in ongoing conversations about performance and career development. Similarly, every manager I have asked will attest to the importance of providing employee feedback and encouraging career growth on an ongoing basis. Yet I have never seen an organization where people are happy with the amount of coaching activity taking place between employees and managers. In sum, the challenge, when it comes to ongoing day-to-day PM behaviors, is not about knowing what to do, it is getting employees and managers to actually do it. It does not advance the field of PM to simply note that managers need to do a better job providing ongoing coaching to employees. What will advance the field of PM is research illustrating what organizational practices increase the frequency and effectiveness of manager coaching behavior. Unfortunately this major challenge to PM is not explored in the article.
Avoiding a Problem Is Not the Same as Fixing It
Most PM methods can be broadly placed into two general categories.
• Employee classification involves assessing employee performance to support decisions about where to invest scarce resources such as pay, promotions, or limited development opportunities (e.g., job assignments, expensive training courses).
• Employee development involves assessing employee performance to provide coaching feedback and advice to increase effectiveness.
Both categories require evaluating employee job performance, but how employees should be evaluated is different, depending on whether the focus is on classification or development. Classification involves comparing employees against one another to determine which employees deserve higher pay raises, development resources, or promotion opportunities. Development evaluations may provide qualitative descriptions of employee performance with no comparative information at all. Assessments that stress development tend to avoid normative evaluations like ratings that directly compare people against one another because such normative evaluations of performance can hurt development (Dweck, Reference Dweck and Dienstbier1990; Roch, Sternburgh, & Caputo, Reference Roch, Sternburgh and Caputo2007). But these sorts of descriptive, developmental evaluations will not help companies that are seeking to create fair, consistent, and accurate methods to place employees into performance categories for the purpose of compensation, development, or staffing. It does not make sense to argue whether classification is more or less important than development is. Companies must both evaluate and develop employees to create a high performance workforce. The key is to build PM processes that effectively balance both needs.
In my experience, the single most difficult aspect of PM is creating methods that classify employees on the basis of performance in a manner that is considered accurate by company leaders and perceived as fair by employees. This requires dealing with the reality that some employees perform at a higher level than others do. PM would be much easier if everyone performed at the same level or if people never felt insulted or threatened by critical performance reviews. But all people do not perform at the same level, and it is important to differentiate between employees on the basis of their relative contributions. Similarly, people do react emotionally to performance evaluations, and it is important to ensure that employees do not feel like “losers” just because they received a lower performance rating than some of their peers did. This requires a mixture of critical yet supportive feedback on past performance, constructive suggestions for increasing future performance, and clear consequences and rewards associated with long-term performance accomplishments.
Rather than addressing the challenges associated with those aspects of PM that involve classifying employees, Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) appear to have narrowly redefined PM so all it involves is development. The PM process they have described does not include methods for dealing with poor performance or for identifying and rewarding high performers. As a result, they have largely ignored the most challenging side of PM. In reality, this side of PM cannot be ignored. If a company adjusts pay each year so some employees receive more than others do, then the company conducts an annual performance review, even if it does not call it an annual performance review. I have worked with companies where employees and managers basically view the compensation process as the performance evaluation process, even though many things that are unrelated to actual employee performance may influence compensation decisions. In my experience, most companies that are said to have eliminated annual reviews probably have not actually eliminated them. The companies have just hidden annual reviews in the back room so employees no longer understand how those reviews are conducted.
In order to fix PM, it is critical to determine how to conduct classification-focused performance evaluations in a way that maximizes employee performance and retention. This is a different question than how to better support employee development activities. Companies must use both classification and development to create high performance organizations. Recommendations that suggest that companies can ignore or abandon processes used for performance evaluations and just focus on coaching are about as sound as recommendations that suggest that you can maintain a healthy lifestyle solely through exercise while completely ignoring your diet.
Can You Fix Something That Does Not Have a Permanent Solution?
The title of Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) implies that PM can be fixed, but does it make sense to talk about fixing something that does not have a fixed solution? As Pulakos et al. themselves have stated, “there are no specific PM components that should always be retained or eliminated” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 30). The value of PM methods depends on a company's current business needs, its past culture, its employees’ characteristics, its resources, and a range of other variables. Methods that work well for a company at one time in history may not work equally well for another company or at another time.
A good example that illustrates this point is the forced-ranking methods that were once famously used by General Electric and Microsoft but that have since been largely abandoned by both companies. A lot of people criticize forced-ranking methods now, but when they were used in the 1990s they were widely endorsed. Business leaders supported them. Employees may have complained about them, but employees still flocked to these companies in droves. People who held shares of General Electric and Microsoft stock during the 1990s would be hard pressed to complain about the financial results associated with those organizations’ PM methods. One could argue whether forced ranking was a contributor to or detractor from the success of General Electric and Microsoft in the 1990s, but it certainly had the appearance of working for those companies at that time. However, PM methods that once worked well can become damaging in a different business and labor market. In recognition of this, both General Electric and Microsoft have discontinued the forced-ranking methods that were once viewed as cornerstones of their PM philosophy.
It is unlikely that any PM method will be universally and permanently effective across all jobs and companies. Methods that work now can become detrimental as a company's business climate and employee population changes. In light of this, I question the very concept of whether one can fix PM. It is probably more appropriate to talk about optimizing PM methods for any given company at any given point in time.
Conclusion
Lest anyone draw the wrong conclusion from my comments, I want to emphasize that I am thankful to Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) for writing an article that challenges scientists and practitioners to rethink PM. When companies list human resource processes that fail to meet expectations, PM is almost always included. I also appreciate and agree with many of the observations Pulakos et al. have made about factors that undermine PM effectiveness. My concerns are mainly around how Pulakos et al. have presented their observations. On the positive side, Pulakos et al. have clearly stated that there is no one best way to do PM, but they have also made a lot of claims and criticisms about PM methods that are not always well supported.
PM is a critical yet difficult part of talent management. PM is difficult because few companies feel they do it well. It is critical because companies continue to implement PM even though they do not necessarily like it. What the field of PM needs is more well-balanced, thoughtful discussion and research into the advantages and limitations of different PM methods. What the field does not need are broad sweeping statements suggesting that one PM practice be totally abandoned or that another practice be universally endorsed. PM methods have a significant and direct impact on the lives and employment of employees. It is important to be cautious when making recommendations about PM. There is a strong need for more systematic, unbiased research into questions such as the following: (a) How can companies get managers to proactively engage employees in ongoing performance feedback and coaching sessions? (b) What is the best way to communicate pay and promotion decisions so employees feel motivated as opposed to disheartened? (c) What are the boundary conditions that determine when different types of PM methods are likely to be effective?
I am glad Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have started this conversation, even if I do not agree with all of what they have said. I hope the conversation continues and draws in some intrepid researchers who will empirically study when and why certain PM methods do or do not work. Regarding whether it is possible to actually fix PM, I end with the following criteria that I often use when talking with companies about the efficacy of their PM methods: “You can say you have a great PM process when your company leaders can quickly reach consensus on who the top 20% most valuable employees are in the organization and can explain their decision to the other 80% of employees in a manner that inspires them to improve their performance and does not lead them to give up hope, quit, or call their lawyers.”
Since 2007, my work has involved helping companies to maximize the value of a technology system that supports strategic human resource processes, including performance management (PM). This technology is largely agnostic when it comes to PM process design. It can be configured for PM processes that make extensive ratings or avoid ratings entirely, encourage elaborate goal cascading or completely forego the use of goals, focus on once a year PM events or encourage ongoing employee–manager check-ins, and so forth. In sum, the system can be configured in any number of ways.
The conversations I have with companies focus on determining what type of PM process will be the most effective in their particular organization. This will ensure the technology is appropriately configured to meet their unique needs. As a result of my role, I have discussed PM concepts with hundreds of companies crossing virtually every industry and located in almost every region of the world. Most of the companies I work with have thousands of employees, although I have worked with organizations with as few as 50 employees.
Given what I do for a living, I was quite interested to read the article by Pulakos, Mueller Hanson, Arad, and Moye (2015). My reaction to this article is mixed. On one hand, I applaud their efforts to explore new methods to increase the value of PM. On the other hand, I felt the article has failed to clearly define and systematically address the real challenges that impact PM effectiveness. It also criticized many common PM methods without acknowledging the value these methods provide for a great many organizations. The result is a highly skewed interpretation of what PM is, why it is difficult, and how to fix it. Nevertheless, the article does generate attention in an area that is in dire need of more research. So even if I do not always agree with how Pulakos et al. have discussed the topic of PM process improvement, I am glad they have generated the discussion
My overriding concern with Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) is a tendency to make sweeping statements suggesting that certain PM methods are universally good or bad. To their credit, they have stated that “there are no specific PM components that should always be retained or eliminated . . . they will depend on the way that PM system information is currently used; its value; and the specific strategy, PM goals, and needs of the given organization” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 30). But this statement appears after dozens of unqualified or weakly constrained comments about the effectiveness of different PM methods, such as “cascading goals are a significant time sink” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 10), “rating processes have grown overly complex” (p. 11), or “ratings add little to decision making” (p. 16). Many statements in the article suggest that the “best” way to do PM is to focus on coaching, whereas others suggest that use of certain PM methods related to goal setting and ratings are almost always doomed.
Observations made working with hundreds of companies to improve PM lead me to believe that there is no single way to “fix” PM because methods that work very well in one company may not work in another. For example, I have seen rigorous goal cascading methods that work very effectively in established retail organizations but that would probably fail miserably in high-tech start-up organizations. The best approach depends on the business needs and culture of the company, the nature of the employees and the jobs they are performing, and the skills and incentives given to managers. Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have acknowledged this, but they have not defined the criteria that influence when different PM methods are likely to succeed. Instead, they have presented several interesting PM methods that probably work very well in some companies but that I suspect would fail in others. For example, they have suggested using PM methods that emphasize giving employees a sense of autonomy, mastery, and purpose while downplaying ties to tangible decisions about employee pay and promotions. Disconnecting PM processes from pay decisions might work in some companies, but it could have a devastating impact in companies where employees are concerned about procedural justice regarding decisions that impact their financial stability (Colquitt, Reference Colquitt2001). In sum, although this article is full of “good ideas” for improving PM, what is really needed is more clarity about and research into the specific conditions in which these different good ideas actually work in practice.
Where Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) Seem To Be “Missing the Mark”
There are many places in the Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) article where I found myself thinking, “yes, that is true in some situations, but it is certainly not true for all companies or jobs.” The following are some of the major areas about which I believe Pulakos et al. are only “half right.”
Performance Management Means Different Things to Different Companies
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have noted that “the promise of PM can only be realized if there is a shared mindset about what effective PM is” (p. 17). Yet Pulakos et al. have never actually defined what the term “performance management” means. For the purpose of this commentary, I define PM as
standardized and defined processes used to communicate job expectations to employees, evaluate employees against those expectations, and utilize these evaluations to guide talent management decisions related to compensation, staffing and development. (Hunt, Reference Hunt2014, p. 151)
Even this definition of PM can be interpreted in a lot of ways. I have worked with companies that think of PM narrowly, as a method primarily used to guide compensation decisions, and I have worked with companies that think of it broadly, as a process to develop and engage high performing employees.
Listed below are four common business objectives often associated with PM:
• complying with legal or contractual requirements that require clearly defining and communicating employee performance expectations and consistently evaluating and documenting employee performance on the basis of these expectations;
• engaging high performers by identifying employees who are making the strongest contributions to the organization and providing them with significantly greater rewards and recognition in the form of pay, career opportunities, and development resources;
• increasing employee performance and retention through enabling greater levels of coaching, constructive feedback, and dialogue around career development; and
• establishing consistent, well-defined processes and job-relevant criteria to guide pay, promotion, and termination actions.
All these objectives are valid reasons for implementing a PM process. However, a PM process that supports one of these objectives might not effectively support another. For example, the PM methods that Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have proposed would probably effectively support only one of these four objectives. So the suggestion that the methods that Pulakos et al. have outlined will fix PM is clearly limited to a fairly narrow definition of what PM involves.
PM Processes That Fail in One Company Can Be Highly Successful in Another
Much of the evidence Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have provided for the validity of their PM methods is drawn from a single company with a unique culture and set of business requirements. The article is largely a case study describing how this one company is seeking to change its PM process. Such case studies are useful because one can learn from what others are trying. But it is premature to state whether the PM methods advocated by Pulakos et al. will generalize to other companies.
Nor should we accept that Pulakos et al.'s (2015) criticism of different PM techniques will generalize to all other companies. To the contrary, many companies gain significant value from PM methods that Pulakos et al. have described with extremely negative terms such as “time sinks,” “detracting from performance,” or “overly complex.” For example, the evidence in Table 1 provides a marked contrast to many of the negative comments Pulakos et al. have made regarding the value of goal cascading, performance ratings, and annual goal setting. Because Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have justified the value of their PM process design recommendations using qualitative comments from managers (p. 28), I have used the same method here. I am not at liberty to share the names of the companies from which I have drawn these comments, but they are all direct quotes from line managers and employees using these PM methods.
Table 1 Manager Reactions to Performance Management Methods
The point is not to argue about whether one PM method is better or worse than another is. It is merely to note that many assertions that Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have made about the value of certain PM methods is not reflective of experiences found in other companies. Furthermore, the use of more evaluative PM methods such as ratings and pay for performance, which Pulakos et al. seem to have suggested are not worthwhile, has been shown to positively correlate with company profit and growth (Bloom & Van Reenen, Reference Bloom and Van Reenen2007). These more evaluative PM methods may be harder to implement, and managers and employees may not always like these methods, but many things that people find difficult and do not enjoy are often good for them.
Pay for Performance Is Very Effective if It Is Done Correctly in the Right Setting
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have cited the work of Daniel Pink in claiming that “carrot-and-stick rewards are not effective for driving performance or behavior in today's complex work environment” (p. 18). Aside from the questionable generalizability of Pink's theories of motivation across different jobs and employee populations (Hebert, Reference Hebert2010), this claim ignores research that shows pay for performance techniques to be very effective if applied appropriately in many settings (Rynes, Gerhart, & Parks, Reference Rynes, Gerhart and Parks2005). In sum, pay for performance, like most PM methods, may not work everywhere, but it can work very well in certain settings when used the right way.
There Is Nothing Inherently Wrong With Annually Based PM Methods
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have made several statements suggesting that annual PM processes are somehow inferior to methods that do not follow an annual cycle. For example, Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have stated the following: “the yearly review that provides backward-looking feedback is not only ineffective but also potentially damaging” (p. 16), “rather than set goals once per year . . . it is more effective to set ongoing expectations as needed” (p. 20), and “setting SMART objectives annually . . . ignores the dynamic and ever-changing nature of work” (p. 20). I agree with the elements of these statements that call out the value of forward-focused performance discussions and the need to update goals throughout the year to reflect changing business conditions. However, I disagree with the implication that using annual methods is an intrinsic part of the problem. For example, why should one expect that annual PM reviews are inherently backward looking and that ongoing PM feedback conversations are more likely to be forward looking? I suspect managers providing feedback in real time are very likely to be backward looking, calling attention to employee's recent mistakes, rather than stopping and planning out a forward-looking conversation. Similarly, because many corporate goals are set annually, it makes logical sense to establish initial employee goals annually with the same calendar as long as companies recognize that these employee goals are likely to be modified to reflect tactical and strategic changes that occur throughout the year.
Any PM activity can be effective or ineffective regardless of whether it is an annual review or a weekly check-in. The effectiveness of things like goal setting and PM reviews is a function of how well managers and employees carry these tasks out. It is not a function of whether companies choose to link these tasks to an annual calendar.
Pointing Out What Companies “Should Do” Will Not Fix the Core Problems With PM
Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have emphasized that much of the value of PM lies in PM's ability to create better conversations about performance between employees and managers and that these conversations tend to happen outside of PM forms and process steps. I agree that this point is true, but it is not a particularly new observation, nor is it very helpful to companies that are seeking to create more ongoing performance discussions. Every PM training program I have seen emphasizes the importance of not treating PM as a once a year process. PM training typically includes statements like “there should be no surprises during a PM review” and encourages managers to engage employees in ongoing conversations about performance and career development. Similarly, every manager I have asked will attest to the importance of providing employee feedback and encouraging career growth on an ongoing basis. Yet I have never seen an organization where people are happy with the amount of coaching activity taking place between employees and managers. In sum, the challenge, when it comes to ongoing day-to-day PM behaviors, is not about knowing what to do, it is getting employees and managers to actually do it. It does not advance the field of PM to simply note that managers need to do a better job providing ongoing coaching to employees. What will advance the field of PM is research illustrating what organizational practices increase the frequency and effectiveness of manager coaching behavior. Unfortunately this major challenge to PM is not explored in the article.
Avoiding a Problem Is Not the Same as Fixing It
Most PM methods can be broadly placed into two general categories.
• Employee classification involves assessing employee performance to support decisions about where to invest scarce resources such as pay, promotions, or limited development opportunities (e.g., job assignments, expensive training courses).
• Employee development involves assessing employee performance to provide coaching feedback and advice to increase effectiveness.
Both categories require evaluating employee job performance, but how employees should be evaluated is different, depending on whether the focus is on classification or development. Classification involves comparing employees against one another to determine which employees deserve higher pay raises, development resources, or promotion opportunities. Development evaluations may provide qualitative descriptions of employee performance with no comparative information at all. Assessments that stress development tend to avoid normative evaluations like ratings that directly compare people against one another because such normative evaluations of performance can hurt development (Dweck, Reference Dweck and Dienstbier1990; Roch, Sternburgh, & Caputo, Reference Roch, Sternburgh and Caputo2007). But these sorts of descriptive, developmental evaluations will not help companies that are seeking to create fair, consistent, and accurate methods to place employees into performance categories for the purpose of compensation, development, or staffing. It does not make sense to argue whether classification is more or less important than development is. Companies must both evaluate and develop employees to create a high performance workforce. The key is to build PM processes that effectively balance both needs.
In my experience, the single most difficult aspect of PM is creating methods that classify employees on the basis of performance in a manner that is considered accurate by company leaders and perceived as fair by employees. This requires dealing with the reality that some employees perform at a higher level than others do. PM would be much easier if everyone performed at the same level or if people never felt insulted or threatened by critical performance reviews. But all people do not perform at the same level, and it is important to differentiate between employees on the basis of their relative contributions. Similarly, people do react emotionally to performance evaluations, and it is important to ensure that employees do not feel like “losers” just because they received a lower performance rating than some of their peers did. This requires a mixture of critical yet supportive feedback on past performance, constructive suggestions for increasing future performance, and clear consequences and rewards associated with long-term performance accomplishments.
Rather than addressing the challenges associated with those aspects of PM that involve classifying employees, Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) appear to have narrowly redefined PM so all it involves is development. The PM process they have described does not include methods for dealing with poor performance or for identifying and rewarding high performers. As a result, they have largely ignored the most challenging side of PM. In reality, this side of PM cannot be ignored. If a company adjusts pay each year so some employees receive more than others do, then the company conducts an annual performance review, even if it does not call it an annual performance review. I have worked with companies where employees and managers basically view the compensation process as the performance evaluation process, even though many things that are unrelated to actual employee performance may influence compensation decisions. In my experience, most companies that are said to have eliminated annual reviews probably have not actually eliminated them. The companies have just hidden annual reviews in the back room so employees no longer understand how those reviews are conducted.
In order to fix PM, it is critical to determine how to conduct classification-focused performance evaluations in a way that maximizes employee performance and retention. This is a different question than how to better support employee development activities. Companies must use both classification and development to create high performance organizations. Recommendations that suggest that companies can ignore or abandon processes used for performance evaluations and just focus on coaching are about as sound as recommendations that suggest that you can maintain a healthy lifestyle solely through exercise while completely ignoring your diet.
Can You Fix Something That Does Not Have a Permanent Solution?
The title of Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) implies that PM can be fixed, but does it make sense to talk about fixing something that does not have a fixed solution? As Pulakos et al. themselves have stated, “there are no specific PM components that should always be retained or eliminated” (Pulakos et al., Reference Pulakos, Hanson, Arad and Moye2015, p. 30). The value of PM methods depends on a company's current business needs, its past culture, its employees’ characteristics, its resources, and a range of other variables. Methods that work well for a company at one time in history may not work equally well for another company or at another time.
A good example that illustrates this point is the forced-ranking methods that were once famously used by General Electric and Microsoft but that have since been largely abandoned by both companies. A lot of people criticize forced-ranking methods now, but when they were used in the 1990s they were widely endorsed. Business leaders supported them. Employees may have complained about them, but employees still flocked to these companies in droves. People who held shares of General Electric and Microsoft stock during the 1990s would be hard pressed to complain about the financial results associated with those organizations’ PM methods. One could argue whether forced ranking was a contributor to or detractor from the success of General Electric and Microsoft in the 1990s, but it certainly had the appearance of working for those companies at that time. However, PM methods that once worked well can become damaging in a different business and labor market. In recognition of this, both General Electric and Microsoft have discontinued the forced-ranking methods that were once viewed as cornerstones of their PM philosophy.
It is unlikely that any PM method will be universally and permanently effective across all jobs and companies. Methods that work now can become detrimental as a company's business climate and employee population changes. In light of this, I question the very concept of whether one can fix PM. It is probably more appropriate to talk about optimizing PM methods for any given company at any given point in time.
Conclusion
Lest anyone draw the wrong conclusion from my comments, I want to emphasize that I am thankful to Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) for writing an article that challenges scientists and practitioners to rethink PM. When companies list human resource processes that fail to meet expectations, PM is almost always included. I also appreciate and agree with many of the observations Pulakos et al. have made about factors that undermine PM effectiveness. My concerns are mainly around how Pulakos et al. have presented their observations. On the positive side, Pulakos et al. have clearly stated that there is no one best way to do PM, but they have also made a lot of claims and criticisms about PM methods that are not always well supported.
PM is a critical yet difficult part of talent management. PM is difficult because few companies feel they do it well. It is critical because companies continue to implement PM even though they do not necessarily like it. What the field of PM needs is more well-balanced, thoughtful discussion and research into the advantages and limitations of different PM methods. What the field does not need are broad sweeping statements suggesting that one PM practice be totally abandoned or that another practice be universally endorsed. PM methods have a significant and direct impact on the lives and employment of employees. It is important to be cautious when making recommendations about PM. There is a strong need for more systematic, unbiased research into questions such as the following: (a) How can companies get managers to proactively engage employees in ongoing performance feedback and coaching sessions? (b) What is the best way to communicate pay and promotion decisions so employees feel motivated as opposed to disheartened? (c) What are the boundary conditions that determine when different types of PM methods are likely to be effective?
I am glad Pulakos et al. (Reference Pulakos, Hanson, Arad and Moye2015) have started this conversation, even if I do not agree with all of what they have said. I hope the conversation continues and draws in some intrepid researchers who will empirically study when and why certain PM methods do or do not work. Regarding whether it is possible to actually fix PM, I end with the following criteria that I often use when talking with companies about the efficacy of their PM methods: “You can say you have a great PM process when your company leaders can quickly reach consensus on who the top 20% most valuable employees are in the organization and can explain their decision to the other 80% of employees in a manner that inspires them to improve their performance and does not lead them to give up hope, quit, or call their lawyers.”