Wednesday, April 16, 2008

Can Performance be Measured?

John Fletcher posted an interesting article "Why Quantitative Measures Often Make Performance Worse, not Better" on the Slow Leadership Blog about how quantifiable objectives is more about office politics than performance.

The article takes the point of view that performance measurement does not work well because typically what is measured is not what matters the most, and that performance often delines when it is measured because once people reach their target there is no incentive to exceed it.

I previously shared some information on performance measurement as well as some personal stories. I agree with John that sometimes quantifiable objectives can be more about office politics and that they can have undesirable effects and he raises a several good points. However, I will say that performance cannot be evaluated without well defined criteria. The article also seems to say that any quantifiable objectives are not as good as less tangible measures, a statement which I can't agree with.

The example of setting set targets also shows what to avoid; In the example, the goal is to answer 90% of inquiries - answer less than 90% and the employee does not get a bonus, answer more than 90% and the employee receives a bonus. This strategy could lead to employees doing the bare minimum of work to reach 90%, but not strive to achieve 100%. If the department's goal was to answer as many queries as possible, a better solution would be to use a quota approach: achieve target and receive a bonus, answer between 90 and 95% of inquiries and receive a bigger bonus, answer between 96 and 99% of inquires and receive an even bigger bonus, and answer 100% of inquiries and get the max bonus. This way, at least in theory, employees will always be motivated to exceed the target.

Choosing "good" performance measures and metrics is one of the most critical aspects of designing an incentive plan, while"bad" measures could result in encouraging undesired behaviors. It is important to know exactly what the desired behaviors and goals are before choosing any measurement. Also, as David Cichelli from The Alexander Group and Liz Cobb from Makana Solutions pointed out earlier, too many measures can ruin a compensation plan by making it overly complex and confusing the payees.

1 comment:

Anonymous said...

I'm not against measuring outcomes - but they do pose a problem when those measures are part of an incentive. It is my experience that measuring behaviors that lead to outcomes is a better way to structure incentives.

The key is to find those behaviors that are vital to the outcome desired. Reward the behavior - not the outcome.

I also try to find a way around percentage measures. Percentages allow for manipulation. There needs to be a combination of percentage and real number.

In most of the programs I deal with we try to combine some sort of calculation that takes into account percentages and "units."

Blog Search

Blog Directories

Sales Blogs - BlogCatalog Blog Directory

Enter your email address:

Delivered by FeedBurner

Add to Technorati Favorites

Companies Linking to Me


About Me

My photo
Ottawa, Ontario, Canada
Julien Dionne is a well-rounded consultant with global business management experience and outstanding technical, business and leadership skills. He earned a Bachelor of Applied Science in Software Engineering from the University of Ottawa, Canada, and he is a member of the Canadian Professional Sales Association. The views posted within this blog do not reflect the views of Julien’s current or previous employers and clients. Julien can be reached at