Stack-ranking out now: How Microsoft lost a decade

Stack-ranking or forced rating or bell-curving are some of the pet words of HR leaders and Business leaders especially closer to appraisal time. Jack Welch used it to make GE competitive and fit to outperform their own benchmarks! GE had become fat and lazy and had businesses diversified from washing machines to Jet engines. Forced ranking or his 20/70/10 "differentiation" became mantra for weeding out bottom 10 percent of the stack.
The bell-curve ranking system was popularized by GE Chief Executive Jack Welch back in the 1980s being used by many companies, but some of them have dropped it in the last years, including Adobe and Expedia. The main idea of the system is that a company’s employees performance follows a bell-curve, having a small percentage of high performers, a large number of average performers and, at the lower end, another small percentage of underperformers. The system would be used to find and get rid of underperformers in order to improve the overall performance of the workforce and resulting in a smaller standard deviation which corresponds to a more homogeneous company.

Pressure to perform, if maintained below a certain level, can lead to higher performance. However, as the lay-offs take place, constant pressure demoralizes employees, leading to drop in performance. As the bottom performers depart, the rigid distribution of the bell-curve forces managers to categorize a high performer (when compared to the rest of the labor market) as a mediocre. A high performer, unmotivated by such artificial demotion, behaves like a mediocre. Further, in a shrinking company, managers find it difficult to differentiate employees. As a result, they begin to reward visible performance over the actual. Beyond a certain point, the erosion of social capital has the potential to cripple the company.

As the Seattle Times reported, Lisa Brummel, Microsoft’s head of human resources, wrote in a Nov. 12 e-mail to employees: "No more curve. We will continue to invest in a generous rewards budget, but there will no longer be a pre-determined targeted distribution. Managers and leaders will have flexibility to allocate rewards in the manner that best reflects the performance of their teams and individuals, as long as they stay within their compensation budget."

The stack ranking system was put in place for Microsoft's 2011 performance reviews. Employees received one rating based on their manager's assessment and a calibration process (typically with managers one or two levels above the employee). The rating ranged from 1 (highest) to 5 (lowest). 

A 3-rating equaled target compensation that was competitive with what was offered in the local job market. Targeted distribution limits the percentage of workers in each rating category.

Stack ranking is still championed at many companies, and in particular at technology firms. Nevertheless, critics refer to it derisively as "rank and yank," meaning that employees who end up in the bottom ranks are often fired or encouraged to leave.

"If you are hiring competent people, what does it say about an organization when 10 percent of those people fail each year?" management consultant Aubrey Daniels commented to SHRM Online. "Rank stacking is punishing to the rater and the ratee. It produces an environment where employees compete with each other more than with competitors in the marketplace."

However, in a Wall Street Journal op-ed published in the wake of Microsoft's announcement, former General Electric CEO Jack Welch, credited with popularizing stack ranking while running GE, defended the process—when it’s done right.

 "Most experienced businesspeople know that 'rank and yank' is a media-invented, politicized, sledgehammer of a pejorative that perpetuates a myth about a powerfully effective real practice called (more appropriately) differentiation," Welch wrote.

One criticism of stack ranking specifically, and pay differentiation generally, is that it devalues teamwork. Welch responded: "Nonsense. If you want teamwork, you identify it as a value. Then you evaluate and reward people accordingly. You'll get teamwork; I guarantee it."

According to Cliff Stevenson, senior human capital researcher at the Institute for Corporate Productivity (i4cp), Microsoft is right to axe the policy. 

When i4cp asked companies in 2009 how many of them used stack ranking systems, 49% said yes. By 2011, only 14% used them.

The companies the researchers identified as high performing had already moved away from the system. Only 7% of high-performing companies used stack ranking in 2009, and only 6% used it in 2011.

Every single one of the big Human Capital Management (HCM) software companies, from Oracle to CornerstoneOnDemand to SuccessFactors, uses ratings as the lynch pin of its platform. These ratings—of overall performance, and of the many competencies that apparently combine to create overall performance—are the raw material for everything else that the company does to and for its people. These ratings feed compensation. They guide succession planning. They pinpoint performance gaps and then trigger the necessary training and learning interventions. 

Stack-ranking as tool to keep people constantly compete against other colleagues is a reflection of competitive environment, but that process of competition leaves back many looser, many demotivated, many conspiring and self-promoting, leading to 'save job' culture which is highly safety focused than innovation and performance focused. It is actually cannibalistic. 

Need is to treat performance and reward separately and not mistake sense of fear for drive to perform!