Bell Curve for Performance Management – effective tool or poor #leadership let out?

What is a “Bell Curve”?

A bell curve is the shape of a statistical set of data that results in the shape resembling a “bell”, sometimes known as a “normal distribution” curve. It is a way of displaying data to understand strengths & weaknesses in terms of frequency.  It is used in many situations, not just performance management.

How does Bell Curve performance management work?

If you have 10 people in a team, you rank people compared to others in the team on a set of common objectives or KPIs as part of the performance management process. This data is plotted on a chart with score on one axis and numbers on the other

The logic is, as per the bell-curve rule, you have to classify certain number of employees forcefully under predefined rules.

This is forced distribution.

The (forced ranking) process ignores potential. Most often this process is used in the run-up to change or transition (especially when job losses or downsizing is expected, few companies use it as a stainable management approach

Jack Welch whilst at GE introduced a similar performance management process where the bottom 10% were sacked each year. Many other firms emulate this approach, without developing a supportive and developmental culture alongside.

The Bell Curve, published in 1994, was written by Richard Herrnstein and Charles Murray as a work designed to explain, using empirical statistical analysis, the variations in intelligence in American society many believe the “bell curve approach to performance management was derived from this work. (Welch was earlier)

Risks of “Bell Curve Management”

There is a natural assumption in using this model of performance management and forced ranking, which is that you have employed “sub standard” people. That if you like manager have failed to recruit and develop their own people. For the system says that for every “above average” employee you have, you have an underperforming employee. Of course this may be true, but in forced distribution systems there is no accounting for good and effective management.

Reward weak managers, punish effective leaders

Another side effect of these bell curve performance management systems, is that it gives the ineffective and lazy manager a tool with which to “beat” people with. The tool says there must be a poor performer “…and you are that person”, whereas the effective leader, manages, coaches and support the entire team well. Develops or removes poor performers, means that the effective leader needs to mark one or more of their team as “underperforming”

Not statistically valid

While in theory the “bell curve” is a scientifically valid approach, there is one “fact” that gets over looked by many proponents of this approach…. Numbers..

For any statistic to be valid, there needs to be a minimum set of data. The bigger the data the better, however for surveys etc it is often accepted that a base of less than 100 is meaningless. So….

If you are asked to rate your 10 people on a scale of 1-5 (where 3 is average, 1&2 underperforming) you will have:

1 – 1 person

2 – 2 people

3 – 5 people

4 – 2 people

5 – 1 person

Now of course this is replicated across all departments or teams. So what if you have a low performing team of manager A, and a high performing team of manager B? Both have 1 high performer, and 1 low performer. What if the high performer from manager A’s team was no better than average from manage B’s team? Why should manager B and their team member who is the lowest in that team be punished?

Employee Engagement

One of the goals of many organizations is employee engagement. How can an employee that has been rated as “poor” but knows they have been delivering above average in the company feel engaged?

If you want to engage with employees, use engagement based strategies, if you want transition, then use transition based strategies, but do not expect to win at both!

As a discussion tool or in transition or significant change, I believe this is a great approach, but for reward or dismissal, I believe it is a flawed approach that should never be used.