There are a lot of contexts in which you could ask that question.  In our case, we have recently been asking ourselves this question in the context of video storage.  Unlike traditional DVR solutions, there are two types of video storage in our MVaaS architecture - storage on the customer site on the recorder, and centralized storage in the network (actually in a data center attached to the network if you want to pick nits)

The question we are asking ourselves is not how much video is enough to save on the customers premise.  The customer premise question is too straight-forward.  If a customer wants more storage, you put in a bigger hard drive, you give them more storage.  It isn’t challenging technically, it is pretty transparent economically, and it isn’t particularly complicated to think through.  The question we are asking is how much video storage is enough in the network.

This is a bit more complicated thing to think about.  The issue is this - with our MVaaS solution we can enable users to store video off of the recorder at their remote location and into a secure data center.  Users can do this manually by finding video that is interesting and then saving it into the network (so it persists, they can tag/annotate it, share it with others, etc)  They will also be able to do this programmatically (we are about a month away from this) by setting up business rules that will cause video to be pulled off of the recorders and into the network based on specific transactions (like a void transaction on the POS) or events (like the back door opening after hours)  Early customer feedback on this capability has been positive.  The question is how much centralized storage should we include with our service.

Start with the assumption that we are going to provide some amount of storage greater than zero as part of our base packages.  Philosophy here is simply that we want users to try the capability, use it, get hooked and then want more of it - at which point we can charge those users that really use the service while still letting everyone get a taste of the capability.  Add the other assumption that giving away unlimited storage as part of the base package is not a good idea - having a fixed zero price point coupled with an unbounded potential cost structure doesn’t seem too smart and the baking in an average price based on the expected average usage of storage will likely make the base service too expensive for those that don’t use the central storage.

So I’ve narrowed it down then, we are going to include some amount of central storage with our service that is greater than zero but less than infinity.  Isn’t that helpful?

I’m going to violate my self-imposed ban on string-posts and give some more insight tomorrow.  I spent last weekend and the first couple days this week on a vacation with my family (at Legoland among other places) and am still trying to catch up and ditch the nausea I picked up from the Age 5 appropriate Bionacle Blaster ride!  I can definitively say that once is enough when it comes to that ride.