Candy, Flowers and Service Credits
Imagine that on the way to the airport your car’s engine breaks down and you are stuck in the middle of nowhere. Before the cab company can arrange for help you have missed your flight. You get a free ride home and while you ponder over missing that important flight you receive candy, flowers and a voucher for a free ride. The analogy may not be perfect, but service credits are like candy, flowers or vouchers when a service provider fails on something important. I personally dislike the idea of using service credits to compensate for failures. The word penalty sounds better when used in the context of failures but in the absence of provisions to help the client overcome consequences of failures, it is better semantics only.
Speaking of provisions to compensate for failures, one important question comes into fray - whose fault is it anyways? If the cab company was hiring the cars from a rental company, we can see the logic behind the phrase "beyond our control" that most providers like to include in service level agreements. Providers do not see it fair to expect them to take responsibility in arrangements complicated by multiple players and incidents due to acts of God. On the other hand, service clients wish to see owning up for failure as an important part of running software operations for them. And it is not hard to brew the feeling that by throwing service credits vendors (who are rapidly transforming to service providers) have found a way out of avoiding onerous responsibilities to prevent or compensate for failure. However, to make a neutral observation, it must be said that establishing accountability will remain a formidable challenge in contemporary service management. Hopefully, significant strides will be made to do so as providing and utilizing services through the cloud mature and reach mainstream levels. Perhaps new models that entail 3rd party monitoring, mediation, etc will be needed to address this contentious aspect in a way acceptable to both providers and clients.
Currently it is better for clients to go with the realization that service credits are not substantive enough to make up for losses due to failures nor are they deterrent enough to make sloppy providers to put their act together. In simpler words, it is better to prevent failure by choosing providers who are more likely to provide robust services rather than those who are lavish at doling out service credits. There are no studies that indicate any negative correlation between the quality of service provided and generosity with credits, and it is possible that providers who are generous with credits also offer more reliable services, but it is good as a client to be more focused on reliability and to not mix the two. Track record is a good indicator to go by when choosing a provider, but many providers have not been in the market long enough to establish a track record. It is not hard to look for strengths or suitability. Vendors who are serious about uninterrupted service usually make it a point to let clients know what measures they take to the effect. With a few visits, discussions, etc, it should not be hard to sniff out the ones who are serious about their business from the ones who are merely bent on riding the cloud. Providers who have invested in putting good fail-over mechanisms, availability measures and customer support systems are likely to provide problem-free services and rectify problems sooner and more effectively when they occur. So are providers who actually review their services periodically. And good response is the next best thing to not having interruptions.