The final toll: Will the bell curve be retired?

Sanjeev manages one of the most high performing operations teams in one of India’s most reputed knowledge service companies. “This year has been good,” thinks Sanjeev, “because my team had the highest quality scores and we had the biggest output.” One thing has been bothering Sanjeev though! Despite clear evidence that his team has been the best among all other operations teams, Sanjeev’s HR partner reached out to him a couple of days back with the request that he cannot promote more than two. “Be careful Sanjeev,” advised the HR partner, “do not rate all your folks high. Your post-appraisal conversations will be more difficult to handle than you can imagine.” Sanjeev is a typical example of a manager who is caught in what we may call ‘the bell trap.’ After all, since things in life typically follow normal distributions so should performance.

The talent management team in Sanjeev’s organization feels that a bell curve indicates a fair and well-executed performance management strategy. Unfortunately, with special instructions about how to place people’s performance on the bell curve, the talent management team is only trying to game the system and forcibly comply with a belief. But Sanjeev’s company is not the only one. For most organizations, performance management often equates the performance management process or a set of annual activities culminating into the performance appraisal. Of course, as an outcome to the ‘process’ of performance management, employee performance is plotted as per a bell curve. Before we discuss any further, it would be worthwhile to understand some of the common myths that surround performance management.

The first myth among most organizations is that performance management is the same as performance appraisal. In a typical performance management ‘cycle’, an organization conducts a set of activities at specified intervals such as mid-year reviews, training programs, annual appraisal and end-of-year manager-staff conversations. While it is difficult to trace how an organization exactly falls into the ‘appraisal’ trap, it is not difficult to conclude that performance management is grossly ineffective and inadequate in organizations where it is positioned as a process.

A recent survey by Corporate Executive Board indicates that 95 per cent of managers are dissatisfied with their performance management process. The most common notion among employees and staff about performance management processes in organizations is that while HR views the compliance to process as indicative of success, the process in itself is useless unless there is accountability for the use and effectiveness of it. Managers feel that they lose sight of the purpose of performance management in the labyrinth of its mechanics. In terms of process elements, most agree that for performance to be effective, the only real components that matter include coaching, recognition, real-time feedback and solid goal setting. How many performance management processes can claim to be effective in these?

The second myth surrounding performance management is about its outcomes. Most organizations and HR leaders believe that the success of a performance management process is indicated through the profile of a bell-curve. Bell-curve loyalists comment that it is the natural order of things. It is even truer for larger populations and as the sample size of employees included in an annual performance rating process continues to grow, it is natural to expect the shape of the distribution to look like a bell irrespective of the type of company, its demographic distribution, the location and the industry.

Subeer Bakshi, Consulting Leader, Talent & Rewards at Towers Watson has an interesting observation to share. “Let us assume that a company has office entry time of 0900 AM, and the company tracks the entry log of employees in five cohorts of 30 minutes starting at 0730 AM and ending at 1000 AM. Let us also assume that arriving early is somehow an indicator of job performance. In such a scenario, the distribution of people entering the office will follow a bell-curve with most reporting for duty around 9:00 AM. However, if we tracked this by minutes and seconds and not by blocks of time (which is a surrogate for a rating scale) we will inevitably see the long tail/pareto principle in evidence.

The concept of the bell curve comes from the Jack Welch era where it was used as an effective tool to weed out poor performers. GE introduced the forced ranking system in the 1980s as a discipline to ensure that underperforming employees do not stay in the system for too long. In many ways, we may consider GE to be the pioneer behind making the bell curve mainstream.

Let us try and investigate what makes the bell curve so popular and what it does. The bell curve provides a way of rationalizing the difference in performance across the organization under the same internal and external conditions. It also provides a basis to question the validity and authenticity of performance assessments that managers provide of their teams. If a manager rates a majority of her team high or a majority of them low, a bell curve provides the basis for questioning such skews. While ensuring discipline around processes of managing the performance of the organization, it has also come to represent the biggest evil existent in the modern day organization—the bell curve typifying the idea of effective organizational performance management.

While normalization is a natural order of things, one may also consider it as the “average” order of things. For an organization that has recorded “average” performance, a bell curve will indeed turn out to be an accurate representation of the performance of its population. Skews, however, are likely to occur when situations are “off-average”. An example of an off-average business situation is when it has met more than 140 or 150 per cent of its targets. In such a situation, force fitting a majority of the team in the “average” curve is inaccurate and unfair. Similarly, the reverse is also true. Reporting performance of the team in the “average” bell curve when the year has clearly indicated underperformance is inaccurate.

What experts and statisticians describe as a natural outcome has transformed into an impassable and rigid indicator of an effective performance management process. Organizations have become so firm in their need to comply with the bell curve profile that the process of appraisal has become stifling for managers and employees alike. Anandorup Ghose, Head of Rewards at Aon Hewitt says, “There is a correlation between business performance, team performance and employee performance. For a business that has met over 150 per cent of its targets, it would be unfair to expect that the performance profile of its employees will appear like a bell curve because there will be more high performers at the lead end of the curve. Similarly, for a low performing organization, the reverse is true. A bell curve is an accurate representation of performance only for an organization whose performance has been average.”

A rigid approach towards bell curve fitment leads to unfortunate consequences. First, managers are mandated to place only a certain percentage in the top performer bracket. Rating and placing people as per a bell curve mandate introduces ambiguity rather than reducing it. A Towers Watson 2013 study reveals that more than half of the employees in organizations in the global survey believe that high performers are not adequately rewarded. A key reason behind the sentiment was the fact that a forced fitment system promotes a tendency among managers to clump performance rankings towards the middle. Most managers also feel unhappy about the fact that they have to choose among favorites. After Microsoft abolished the bell curve, it freed managers from the worry of needing to place people in a certain profile. Now, managers at Microsoft only need to assess impact and feed it into a tool. The tool takes care of how the impact translates into an individual’s compensation and increment.

In many ways, a bell curve can be thought of as a baseline for introducing discipline in an organization’s performance management process. Shanthi Naresh, India Business Leader, Talent Consulting and Information Solutions at Mercer, believes that a bell curve is a good starting point for an organization in its initial stages of talent management maturity. It provides a framework for the organization to differentiate people and their contributions. At the same time, as talent management processes mature, an organization should rely lesser on it as a performance differentiation mechanism, but may continue to use it as a compensation differentiation tool. Naresh says, “If even a portion of the time spent by organizations on rating discussions and normalization can be used for regular discussions between the manager and employee on goals, ways to achieve the same, and career development actions, the performance management system would have delivered its core objective of enhancing organizational capability.”

A central question at hand though is “what performance profile is ideally suited to replace a bell curve?” As talent management matures and an organization gets more effective in managing the employee lifecycle, it is not unusual to imagine that performance distribution in an organization will not follow a typical bell curve. Josh Bersin, in his article, “The Myth of the Bell Curve” argues that there are several problems with simplifying the people problem as per a bell curve. The first is around the fact that by forcing people into small pockets of high performers and low performers, most of the organization’s money and rewards go to the middle of the curve. Secondly, in case of high performing teams, there will still be a small percentage of people ranked as low performers, which is just plain unfair.

Bersin points towards a research conducted by analysts Ernest O’Boyle Jr. and Herman Aguinis among 198 groups comprising teams of researchers, entertainers, politicians and athletes. The results of the study indicate that rather than following the bell curve pattern, performance of groups mostly followed the “power law” distribution. The new distribution model, also known as long tail, indicates that people are not normally distributed, but rather distributed in the following way—a small number of people who are hyper performers, a broad distribution of people who are good performers and a relatively smaller number of people who are low performers. In this model, there is a very small percentage of hyper performers in the organization who are responsible for taking the organization forward. Most people in the organization are actually below the average, but not too far below. Lastly, there are a sizeable smaller percentage of people in a highly talent mature organization who report low performance, if compared to a simple bell curve fitment.

Bakshi supports the viewpoint that a power curve is possibly a more accurate representation of performance in a highly mature talent organization. He also believes that performance and its impact on the organization also differs based on roles. He cites the example of pilots and customer experience management professionals in an airline. In this scenario, beyond a certain threshold of performance, any additional increment in effort of improving performance by a pilot does not directly impact business performance. On the other hand, every incremental improvement in the performance of a customer experience management professional translates into direct business impact. In a power curve model, thus, people with distinctly superior performance can enjoy significantly higher rewards compared to the rest of the organization. Thus, a power curve also rationalizes people investments. After all, people who feel that they are paid fairly are 4.5 times more likely to be engaged than others.

Microsoft and KPMG feature among the league of companies who have taken the brave leap of abolishing the traditional bell curve centric performance management process. But what have these organizations replaced the bell curve with? Bersin’s March 2015 ‘Whatworks’ states, “performance management is not something that high-impact organizations “do,” but is instead a way that work gets done. The assumption that a bell curve needs to be replaced with another ideal profile is wrong. If a talent management team finds the need for such a best-fit profile, maybe it is time for it to reflect and assess how mature its talent management processes are.

When an organization has mature talent management processes, it is safe to assume that the maturity of HR and talent management has nothing to do with the age of an organization. Some of India’s young and fast growing companies are experimenting with ways to manage and enhance performance, without relying on traditional performance management principles. Kevin Freitas, Global Head of Recruiting & Rewards, at the tech company InMobi says, “There is no formal bell curve that the talent management team at InMobi uses. We also do not explicitly communicate high potential status to HiPos because there is no point in actively harboring differentiation.”

Padmaja Alaganandan, Partner and Leader of People and Change at PwC, says, “There are four key elements of performance—Input, process, output and outcomes. The skills and capabilities of people are inputs, and ways of ensuring that they are used adequately for the benefit of a team and an organization refers to the process. Outputs are what an individual and a team achieves at the end of the year. Lastly, outcomes are the long-term improvements that come in an individual, team, or an organization.” While looking at performance management it is, therefore, important to look at the framework holistically rather than focus on only an individual element. Most traditional performance management systems in organization are flawed in that essence.

Lately, organizations have started realizing that performance management is more an outcome of good talent management practices rather than a set of annual activities. The idea behind good performance management cuts across the employee lifecycle. Organizations that have taken the leap beyond the bell curve have faith in strong talent management processes. With the increasing influence of borderless workforces, tech-enabled productivity and collaboration and networking, talent management has to adapt to a more self-governing model—a system that shifts the responsibility of performance from the organization to the individual. Rather than focus on processes, an organization starts channeling its efforts towards managers, careers and role alignment. As management of performance evolves into a different state, perhaps it’s time for the bell’s final toll.