McKinsey Quarterly

The fairness factor in performance management

| Article

The performance-management process at many companies continues to struggle, but not for lack of efforts to make things better. Of the respondents we surveyed recently, two-thirds made at least one major change to their performance-management systems over the 18 months prior to our survey. With growing frequency, human-resources departments are dispensing with unpopular “forced curve” ranking systems, rejiggering relatively undifferentiated compensation regimes, and digging deeply into employee data for clues to what really drives motivation and performance. (For a look at how Microsoft CEO Satya Nadella is innovating with a system that uses hard and soft performance measures to reshape the culture, see “Microsoft’s next act.”)

Yet companies don’t seem to be making much headway. Employees still complain that the feedback they get feels biased or disconnected from their work. Managers still see performance management as a bureaucratic, box-checking exercise. Half of the executives we surveyed told us that their evaluation and feedback systems have no impact on performance—or even have a negative effect. And certain experiments have gone awry: at some companies, eliminating annual performance reviews without a clear replacement, for example, has led employees to complain of feeling adrift without solid feedback—and some employers to reinstate the old review systems.

Amid ongoing dissatisfaction and experimentation, our research suggests that there’s a performance-management issue that’s hiding in plain sight: it’s fairness. In this article, we’ll explain the importance of this fairness factor, describe three priorities for addressing it, and show how technology, when used skillfully, can reinforce a sense of fairness.

The fairness factor

When we speak of fairness, we’re suggesting a tight definition that academics have wrestled with and come to describe as “procedural fairness.”1 It’s far from a platonic ideal but instead addresses, in this context, the practical question of whether employees perceive that central elements of performance management are designed well and function fairly. This eye-of-the-beholder aspect is critical. Our survey research showed that 60 percent of respondents who perceived the performance-management system as fair also stated that it was effective.

More important, the data also crystallized what a fair system looks like. Of course, a host of factors may affect employee perceptions of fairness, but three stood out. Our research suggests that performance-management systems have a much better chance of being perceived as fair when they do these three things:

  1. transparently link employees’ goals to business priorities and maintain a strong element of flexibility
  2. invest in the coaching skills of managers to help them become better arbiters of day-to-day fairness
  3. reward standout performance for some roles, while also managing converging performance for others

Such factors appear to be mutually reinforcing. Among companies that implemented all three, 84 percent of executives reported they had an effective performance-management system. These respondents were 12 times more likely to report positive results than those who said their companies hadn’t implemented any of the three (exhibit).

Our research wasn’t longitudinal, so we can’t say for sure whether fairness has become more important in recent years, but it wouldn’t be surprising if it had. After all, organizations are demanding a lot more from their employees: they expect them to respond quickly to changes in a volatile competitive environment and to be “always on,” agile, and collaborative. As employers’ expectations rise and employees strive to meet them, a heightened desire for recognition and fairness is only natural. And while embattled HR executives and business leaders no doubt want to be fair, fairness is a somewhat vague ideal that demands unpacking.

Winning the battle of perceptions

In working with companies pushing forward on the factors our research highlighted, we have found that these require much greater engagement with employees to help them understand how their efforts matter, a lot more coaching muscle among busy managers, and some delicate recalibration of established compensation systems. Such shifts support a virtuous cycle that helps organizations get down to business on fairness.

1. Linking employees’ goals to business priorities

Building a foundation of trust in performance management means being clear about what you expect from employees and specific about how their work ultimately fits into the larger picture of what the company is trying to accomplish. Contrast that sense of meaning and purpose with the situation at many organizations where the goals of employees are too numerous, too broad, or too prone to irrelevance as events change corporate priorities but the goals of individuals aren’t revisited to reflect them. A typical ground-level reaction: “Managers think we aren’t sophisticated enough to connect the dots, but it’s obvious when our goals get disconnected from what really matters to the company.”

Give employees a say and be flexible. Connecting the dots starts with making employees at all levels feel personally involved in shaping their own goals. Mandating goals from the top down rarely generates the kind of employee engagement companies strive for. At a leading Scandinavian insurer, claims-processing operations were bogged down by surging backlogs, rising costs, and dissatisfied customers and employees. The company formed a working group of executives, managers, and team leaders to define the key areas where it needed to improve. Those sessions served as a blueprint: four overarching goals, linked to the problem areas, could be cascaded down to the key performance indictors (KPIs) at the business-unit and team level and, finally, to the KPIs of individual employees. The KPIs focused on operational measures (such as claims throughput and problem solving on calls), payout measures (like managing contractors and settlement closures), customer satisfaction, and employee morale and retention.

The company took a big further step to get buy-in: it allowed employees to review and provide feedback on the KPIs to assure that these fit their roles. Managers had observed that KPIs needed to vary even for employees in roles with seemingly similar tasks; phone calling for a targeted auto claim is different from skills needed to remedy damage to a factory. So the insurer gave the managers freedom to adjust, collaboratively, the KPIs for different roles while still ensuring a strong degree of consistency. A performance dashboard allowed an employee’s KPIs to be shared openly and daily with team members, making transparent both the teams’ overall progress and the efforts of motivated, top performers.

For the vast majority of traditional roles, this collaborative approach to KPI design is fairly straightforward. For more complex roles and situations—such as when tasks are deeply interdependent across a web of contributors—it can be more challenging to land on objective measurements. Such complex circumstances call for even more frequent feedback and for getting more rigorous about joint alignment on goals.

Adapt goals as often as needed. In today’s business environment, goals set at a high level in the strategy room are often modified in a few months’ time. Yet KPIs down the line are rarely adjusted. While we’re not suggesting that employees’ goals should become moving targets, they should certainly be revised in response to shifting strategies or evolving market conditions. Revisiting goals throughout the year avoids wasted effort by employees and prevents goals from drifting into meaninglessness by year-end, undermining trust. Of respondents who reported that their companies managed performance effectively, 62 percent said that those organizations revisit goals regularly—some on an ad hoc basis, and some twice a year or more. Managers must be on point for this, as we’ll explain next.

2. Teaching your managers to be coaches

Managers are at the proverbial coal face, where the hard work of implementing the performance requirements embodied in KPIs gets done. They also know the most about individual employees, their capabilities, and their development needs. Much of the fairness and fidelity of performance-management procedures therefore rests on the ability of managers to become effective coaches. Less than 30 percent of our survey respondents, however, said that their managers are good coaches. When managers don’t do this well, only 15 percent of respondents reported that the performance-management system was effective.

Start with agility. In a volatile business environment, good coaches master the flux, which means fighting the default position: goal setting at the year’s beginning ends with a perfunctory year-end evaluation that doesn’t match reality. At the Scandinavian insurer, team leaders meet weekly with supervisors to determine whether KPI targets and measures are in sync with current business conditions. If they aren’t, these managers reweight measures as needed given the operating data. Then, in coaching sessions with team members, the managers discuss and adjust goals, empowering everyone. Even when things aren’t in flux, managers have daily check-ins with their teams and do weekly team-performance roundups. They review the work of individual team members monthly. They keep abreast of the specifics of KPI fulfillment, with a dashboard that flashes red for below-average work across KPI components. When employees get two red lights, they receive written feedback and three hours of extra coaching.

Invest in capabilities. The soft skills needed to conduct meaningful performance conversations don’t come naturally to many managers, who often perform poorly in uncomfortable situations. Building their confidence and ability to evaluate performance fairly and to nudge employees to higher levels of achievement are both musts. While the frequency of performance conversations matters, our research emphasizes that their quality has the greatest impact.

One European bank transformed its performance-management system by holding workshops on the art of mastering difficult conversations and giving feedback to employees who are missing the ball. To ready managers for impending steps in the performance-management cycle, the bank requires them to complete skill-validation sessions, moderated by HR, with their peers. Managers receive guidance on how to encourage employees to set multiyear stretch goals that build on their strengths and passions. Just before these goal-setting and development conversations with employees take place, managers and peers scrum it out to test each other’s ideas and refine their messages.

Make it sustainable. At the European bank, the support sessions aren’t one-off exercises; they have become a central element in efforts to build a cadre of strong coaches. That required some organizational rebalancing. In this case, the bank restructured aspects of HR’s role: one key unit now focuses solely on enhancing the capabilities of managers and their impact on the business and is freed up from transactional HR activities. Separate people-services and solutions groups handle HR’s administrative and technical responsibilities. To break through legacy functional mind-sets and help HR directors think strategically, they went through a mandated HR Excellence training program.

The Scandinavian insurance company chose a different road, seeking to disseminate a stronger performance-management culture by training “champions” in specific areas, such as how to set goals aligned with KPIs. These champions then ran “train the trainer” workshops to spread the new coaching practices throughout the organization. Better performance conversations, along with a growing understanding of how and when to coach, increased perceived fairness and employee engagement. Productivity subsequently improved by 15 to 20 percent.

3. Differentiating compensation

Capable coaches with better goal-setting skills should take some of the pain out of aligning compensation—and they do to an extent. However, new organizational roles and performance patterns that skew to top employees add to the challenges. Incentives for traditional sales forces remain pretty intuitive: more effort (measured by client contacts) brings in more revenue and, mostly likely, higher pay. It’s harder to find the right benchmarks or to differentiate among top, middle, and low performers when roles are interdependent, collaboration is critical, and results can’t easily be traced to individual efforts. The only way, in our experience, is to carefully tinker your way to a balanced measurement approach, however challenging that may be. Above all, keep things simple at base, so managers can clearly explain the reasons for a pay decision and employees can understand them. Here are a few principles we’ve seen work:

Don’t kill ratings. In the quest to take the anxiety out of performance management—especially when there’s a bulge of middle-range performers—it is tempting to do away with rating systems. Yet companies that have tried this approach often struggle to help employees know where they stand, why their pay is what it is, what would constitute fair rewards for different levels of performance, and which guidelines underpin incentive structures. Just 16 percent of respondents at companies where compensation wasn’t differentiated deemed the performance-management system effective.

Dampen variations in the middle. With middle-of-the-pack performers working in collaborative team environments, it’s risky for companies to have sizable differences in compensation among team members, because some of them may see these as unfair and unwarranted. Creating the perception that there are “haves” and “have-nots” in the company outweighs any benefit that might be derived from engineering granular pay differences in the name of optimizing performance.

Cirque du Soleil manages this issue by setting, for all employees, a base salary that aligns with market rates. It also reviews labor markets to determine the rate of annual increases that almost all its employees receive. It pays middling performers fairly and consistently across the group, and the differences among such employees tend to be small. Managers have found that this approach has fostered a sense of fairness, while avoiding invidious pay comparisons. Managers can opt not to reward truly low performers. Cirque du Soleil (and others) have also found ways to keep employees in the middle range of performance and responsibilities whose star is on the rise happy: incentives that are not just financial, such as explicit praise, coaching, or special stretch assignments.

Embrace the power curve for standout performers. Research has emerged suggesting that the distribution of performance at most companies follows a “power curve”: 20 percent of employees generate 80 percent of the value. We noted this idea in a previous article on performance management and are starting to see more evidence that companies are embracing it by giving exceptional performers outsized rewards—typically, a premium of at least 15 to 20 percent above what those in the middle get—even as these companies distribute compensation more uniformly across the broad midsection.

At Cirque du Soleil, managers nominate their highest-performing employees and calibrate pay increases and other rewards. Top performers may receive dramatically more than middle and low performers. In our experience, employees in the middle instinctively get the need for differentiation because it’s no secret to them which of their colleagues push the needle furthest. Indeed, we’ve heard rumblings about unfair systems that don’t recognize top performers. (For a counterpoint to radical performance differentiation, see “Digging deep for organizational innovation,” where Hilcorp CEO Greg Lalicker explains how the oil and gas producer sets exacting production standards and then—if they’re met—gives every employee a power-curve bonus.)

Innovate with spot bonuses. Recognizing superior effort during the year can also show that managers are engaged and that the system is responsive. Cirque du Soleil rewards extraordinary contributions to special projects with a payment ranging from 2 to 5 percent of the total salary, along with a letter of recognition. In a recent year, 160 of the company’s 3,500 employees were recognized. Spot bonuses avoid inflating salary programs, since the payments don’t become part of the employee’s compensation base.

Technology’s role

Digital technologies are power tools that can increase the speed and reach of a performance-management transformation while reducing administrative costs. They’re generally effective. Sixty-five percent of respondents from companies that have launched performance-related mobile technologies in the past 18 months said that they had a positive effect on the performance of both employees and companies. A mobile app at one global company we know, for example, makes it easier for managers and employees to record and track goals throughout the year. Employees feel more engaged because they know where they stand. The app also nudges managers to conduct more real-time coaching conversations and to refine goals throughout the year.

Does technology affect perceptions of fairness? That depends on how it’s applied. When app-based systems are geared only to increase the efficiency of a process, not so much. However, when they widen the fact base for gauging individual performance, capture diverse perspectives on it, and offer suggestions for development, they can bolster perceived fairness. We have found that two refinements can help digital tools do a better job.

Sweat the small stuff

In an attempt to move away from a manager-led performance system, German e-commerce company Zalando launched an app that gathered real-time performance and development feedback from a variety of sources. The company tested behavioral “nudges” and fine-tuned elements of the app, such as its scoring scale. Yet it found that the quality of written development feedback was poor, since many employees weren’t accustomed to reviewing one another. The company solved this problem redesigning the app’s interface to elicit a holistic picture of each employee’s strengths and weaknesses, and by posing a direct question about what, specifically, an employee could do to stretch his or her performance. The company also found that feedback tended to be unduly positive: 5 out of 5 became the scoring norm. It did A/B testing on the text describing the rating scale and included a behavioral nudge warning that top scores should be awarded only for exceptional performance, which remedied the grade inflation.

Separate development from evaluative feedback

Digitally enabled, real-time feedback produces a welter of crowdsourced data from colleagues, and so does information streaming from gamified problem-solving apps. The data are powerful, but capturing them can trigger employees’ suspicions that “Big Brother is watching.” One way to address these fears is to distinguish the systems that evaluate employees from those that help them develop. Of course, it is tempting to make all the data gathered through these apps available to an employee’s manager. Yet when employees open themselves to honest feedback from their colleagues about how to do their jobs better, they’re vulnerable—particularly if these development data are fed into evaluation tools. That also undercuts the purpose (and ultimately the benefits) of digitally enabled feedback. Apps should be designed so that employees can decide which feedback they ought to share during their evaluations with managers.

To broaden adoption of the system, Zalando stressed that the app was to be used only for development purposes. That helped spur intense engagement, driving 10,000 users to the app and 60,000 trials in the first few months. Employees reacted positively to sharing and evaluating data that would help them cultivate job strengths. With that base of trust, Zalando designed a performance dashboard where all employees can see, in one place, all the quantitative and qualitative feedback they have received for both development and evaluation. The tool also shows individuals how their feedback compares with that of the average scores on their teams and of people who hold similar jobs.

An agenda of a talent-first CEO

An agenda for the talent-first CEO


The many well-intentioned performance-management experiments now under way run the risk of falling short unless a sense of fairness underpins them. We’ve presented data and examples suggesting why that’s true and how to change perceptions. At the risk of oversimplifying, we’d also suggest that busy leaders striving to improve performance management listen to their employees, who have a pretty good idea about what fair looks like: “Just show us the link between what we do and what the company needs, make sure the boss gives us more coaching, and make it all pay.” In our experience, when leaders understand, address, and communicate about the issues at this level, employees see performance management as fair, and the reform efforts of their companies yield better results.

Explore a career with us