If you pay for something, whether it’s a hamburger, car or IP connectivity, you expect to get what you pay for. If you don’t, you want your money back. That’s why Service Level Agreements (SLAs) drive network connectivity and IT infrastructure purchase decisions and provide an on-going ‘insurance policy’ for enterprises with their service providers.
But a recent article by Carol Wilson at Light Reading – Keeping Customers in the Dark – raised the issue of whether enterprises actually use SLA data and therefore whether service providers still needed to provide it.
This excerpt provides a general view of the story:
“[CTOs from major service providers] seemed to agree that while they have to prove to their customers how well their services perform — particularly as they are moving into the all-packet world and away from the protections of the SONET/SDH realm — those efforts are out of proportion with how often the data actually gets used by customers.”
The discussion was surprising on a number of levels and reminded me of the old saying about a tree falling in the woods with nobody around; or in this case, if a customer’s network connection goes down, but they don’t complain about it, did the outage actually happen?
The tricky part with SLAs is that service providers calculate them and enforce them in different ways (and typically only after a customer makes a detailed request) so it’s hard to compare them. Adding more complexity is that as more people head to cloud services, they are juggling multiple SLAs between a cloud provider, cloud connectivity, etc.
Our view of SLAs is perhaps a little different. We view them not only as a performance metric and external deliverable for our customers, but also as a very important internal diagnostic tool to know how and where we could improve to serve our customers better.
In fact, we not only view SLAs as critical, we stake our business on it by backing everything up with a 100% SLA guarantee and proactively crediting a customer starting with the first second its services are impacted along with proactive calls to them and regular status updates until the issue is resolved. Think of it as a Superior Level of Accountability.
Do those comments about SLAs surprise you? How do you use SLA information? Let me know what you think!
