A few months back, I was on a panel organized by SDForum, the panelists (including me) were asked a question: "have we reached the limits on what can be gained by analytics?" (I am paraphrasing, but it was approximately this.) I was somewhat surprised (I would have been less so if someone had asked about process reengineering, automation etc.) that people even can imagine that we have reached the end of analytics. I answered then that all the "smarter planet" activities that we in IBM talk about are great examples of using analytics to make people's lives better -- better water, better transportation, better government -- and I cannot forsee any limits to us improving people's lives using analytics.
But that got me thinking -- how does one quantify the benefits of analytics. Sure, anecdotes abound. But we are data driven. So where is the data on the benefits of analytics.
Then I came across this paper. (I got the link from the "Big" Big Data report by McKinsey.) Through some analysis, and some self reporting by companies, they have concluded that DDD (data driven decision making) adopting companies have 5 - 6% better top line and bottom-line. So for a $10B company, that means $500M improvement. Think about it. How much would that company spend to get that? Just take a look at the cost of storage (I am being very approximate here), on a cloud, we are looking at around $1M/yr for a petabyte of data. Throw in computation, 10x it, we are looking at $10M/yr. $500M on $10M return.
That's it. Obviously, Big Data pays off! Do not "let's throw away because it is too expensive". Instead "store because the cost of analytics + storage more than pays for itself."