Making a point with SLAs
CEO & Founder of Server Density.
Published on the 4th April, 2013.
SLAs are generally financially irrelevant because they typically cap any compensation at the total spend under the claim period e.g. if you spend $100 a month you can claim up to $100. When your spend is significant then it’s nice to get the money back but it doesn’t necessarily reflect the full cost of any service interruption.
Instead, I see SLAs as useful for two reasons:
To signal commitment to a level of service
The level of SLA a company provides indicates its confidence in its own reliability and infrastructure. Traditionally, Rackspace have a very good reputation in managed hosting for uptime; something reflected by their 100% uptime guarantee. But it’s important to look a little further to see what caveats they have: does it include scheduled maintenance (almost never), how do they define uptime and do they differentiate between public/private network connectivity vs power failures as different parts of infrastructure with different SLAs?
This is quite standard so I think the second use is more important:
To make a point about being unhappy with the service
I generally make SLA claims even for the smallest outages and problems. It takes time and effort for a provider to process them and calculate the claim amount, as well as the actual payment, which all count as business costs. I’d bet that these are measured quite closely and so I use SLA claims as a method of making a point about the level of service – to encourage them to fix it.
You can complain about an outage and speak to management but if you want someone to actually notice the issue, make it cost something!