SLA = Stupid Limiting Agreements

SLAs seem to be the staple for the customer management process for contracts, performance and operations.  The first time I heard the word SLA I was consulting for a Fortune 500 IT company and they needed to have a group of metrics because of the poor service they had been delivering to their banking customers.  I already was a student of the statistics of Shewhart and Deming, meaning I understood the difference between “common” and “special” causes of variation and also understood that having a service level agreement (SLA) didn’t improve the performance of the organization.  I used SPC (statistical process control) to determine the differences in variation.  All basic to improving the system.

The problem . . . I was the only one focused on improving the partnership.  The IT vendor and the customer were focused on the service level and not the system.  The customer wanted penalties and the IT vendor wanted rewards (and to avoid penalties).  The two groups spent an inordinate amount of time dickering over what the rewards and penalties should be and I (working for the IT vendor) was to be sure that the operational definition of the metrics was such that the vendor could not fail.  The slew of waste (manipulation, reward/penalty setting, etc.) between the IT vendor and the customer was astonishing.  No one was interested in working together to improve method or even discuss the validity of the original measures. 

SLAs are no more than targets and create what I believe to be adversarial relationships and distrust, focusing on results not method.  This is no different when the SLAs are internal. I see this between departments and units. “I will get you my work in 2 days or less.”  The problem is the measure is not tied to any customer metric it is all internally focused.  Additionally, the amount of manipulation begins when you hear things like “the clock doesn’t start until I open your request” and they don’t open their email for a week . . . did they really hit the SLA?

A better “systems thinking” way is to understand purpose from a customer perspective, derive measures and then find “new” methods.  This avoids the waste associated with measures that do not matter.  Workers that understand good customer metrics and expectations can be creative in changing method.  Partners (like my Fortune 500 IT company and their customers) can achieve continual or continuous improvement by working together on method, not SLAs.

Tripp Babbitt is a speaker, blogger and consultant to service industry (private and public).  He is focused on exposing the problems of command and control management and the termination of bad service through application of new thinking . . . systems thinking.  Download free Understanding Your Organization as a System and gain knowledge of systems thinking or contact us about our intervention services at [email protected].  Reach him on Twitter at

Share This: