Service level agreements existed long before the term “cloud computing” was coined. We’ve grown accustomed to thinking of SLAs as a critical component in our relationship with service providers and customers. They give detail to the service offering bullet points, set our expectations, and should protect both provider and customer. But as cloud computing becomes more pervasive and the features of cloud services flourish, are SLAs as important in cloud computing as we think?
I’ll admit, I’ve never read the SLA or cloud service agreement before subscribing to a new service for personal use. Never. Not once. Call me naïve, ignorant, or too-trusting. I may have perused the agreements later out of curiosity, but my sign-up is mostly determined by whether the service offers the features I need at a price I am willing to pay. Reputation is certainly important to me as well and weighing professional reviews and customer comments is part of my decision-making.
So the process of acquiring personal cloud services isn’t that different from the way I obtain other, non-cloud services. Cable or satellite TV? The choice is made by evaluating which provider offers the channels I want at a price I’m willing to pay. Ground or next-day delivery? That depends on how quickly I think I need the item and again on whether or not the additional cost is worth it to me.
What happens when I don’t receive the service I expect? Most likely only a little grumbling since I’ve been satisfied with previous experiences or my overall service. In extreme cases, I’ll contact the offending company without a thought of consulting the SLA first. Though it may take perseverance and demanding escalation on my part, the company has almost always rectified unsatisfactory service delivery with at least a mostly satisfactory resolution.
However, businesses cannot afford such a relaxed attitude in evaluating and obtaining cloud services. After all, more than personal inconvenience is at stake. At a minimum, cloud service failures or performance issues impact business productivity or cause temporary customer frustration. Longer or more severe disruptions can easily lead to a significant loss of customers and a substantial loss of revenue.
So businesses moving to the cloud must require more comprehensive service definitions, mutually acceptable service levels, and clear resolution or compensation when what’s being paid for isn’t delivered. Though it may be argued that the protection this stricter stance offers may be largely superficial, this is how we achieve some comfort that we’re being treated fairly and that our risk is somehow minimized. In the absence of a virtual handshake, we must at least have a cloud service guarantee. At this point in time, it is the service level agreement that fulfills this need.
I hinted that expanding features of cloud services may lessen our reliance on SLAs at the beginning of this article. Indeed, providers such as Amazon and Microsoft have become much more transparent and offer dashboards and management consoles that easily relay health information about their environment as well as our specific cloud services. However, these features still fail to meet an important requirement we have today: identification of our compensation when service is not “Normal” or “Good”.
Ultimately, we find SLAs crucial in the selection of our cloud service providers today. However, this perception may be waning. As businesses begin to think of cloud computing like instant-on, always-connected, individuals do. Perhaps only features, cost, and reputation will need to be considered.
Before you call me naïve, think about this: Are businesses demanding service level agreements for more mundane, yet necessary, services like electricity and package delivery?