Skip to content

Death of the SLA

March 20, 2014

Service Level Agreements have become a sacrosanct element of today’s contracting practices. They provide the basis for defining and measuring performance and therefore also for judging failure and associated financial penalties.

However, SLAs have their critics. They are too complex; they often result in a ‘green’ dashboard when service users are complaining about quality; they fail to adjust to shifting needs; they result in an environment of blame and recrimination …. the list goes on.

Today I was reviewing a dynamic on-line performance management system, capable of capturing and reporting on numerous criteria. It is therefore flexible to business needs and priorities, which inevitably change over time and also vary by functional group. With a system like this, do we really need the complexity of the SLA, or can we return to the simplicity of a much smaller number of key performance indicators (KPIs)?

Research suggests that there should be no more than 5 KPIs and ideally less. These are the critical success criteria. But rather than drive performance via these fundamental measures, we decide to introduce dozens (on average more than 20) additional ‘service level criteria’, which introduce confusion and complexity.

With on-demand metrics, the parties could instead have contract terms that specify the mechanisms to review and address problems. “Penalties’ would be based not on specific service levels that might have little or no real value, but instead focus exclusively on the achievement of the KPIs.




  1. Dave Connor permalink

    Tim great o hear you commenting on this issue , as always your views are very close to reality. Service Level agreements have been overtaken by scorecards in the last ten years SLAs never really worked becuase they were typically set by the buyer and accepted by seller with no real consequences. Supplier performance management with meaningful and as you say few KPIs that really measure value delivery has overtaken KPIs I agree that SPM should be built in to the contract with gain and pain consequences and not one sided as is still typical in many industries.
    Biznet are the market leader in cloud score carding and I am delighted to be working with them to drive performance improvements in the oil and gas industry.

  2. Jonathan Farrell permalink

    The main problem with SLAs is that they’re designed and written by lawyers as a legal document rather than as an agreement between two parties. I’ve been assisting organizations for about 20 years with SLAs, and have conducted dozens of seminars on how to put SLAs together. My approach is that SLAs need to be in two sections. The first is the legal section containing all the legalese of a contract, typically no more than about 10 – 15 pages. The second section is the operational section, containing all the information necessary for the operations of the SLA including clear service definitions, equipment covered, locations to be covered, etc. This second section must be flexible to cover the changing requirements of the service recipient during the term of the SLA.

    I also agree with Dave Connor’s comments that there should be no more than about 5 KPIs. You can certainly define more in the SLA but you should only choose about 5 – 7 to measure at any one time, Any more creates an administrative nightmare and may detract from the overall purpose of the SLA.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: