Little Data a Thing?
When talking with a few different people, I’ve heard references to projects pulling in the breadth of data collected – I call it "little data" – basically running the other direction from efforts to capture everything and anything for analysis.
Are you seeing this – or considering this in your application work?
The pull in this direction comes from an abundance of caution on one front, and a desire to not have information on file that could be a problem on another front.
The first is one of security and data protection caution. Not wanting to have information in the systems that could be mis-used or represent a *future* issue in terms of information stored. Information that is ok to have today may not be so much tomorrow. You can see the impact of things like this when you consider healthcare industry regulations, credit processors and protection/compliance guidelines, etc.
The second is the risk posed by capturing and processing the information. With the transactional breaches that have occurred of late, it may be easier to to just not be responsible for that information, rather than risk it. This is a very big decision though and (I think) one that can put you at a competitive risk, depending on the information you decide not to capture or process with your own systems. Some companies are minimalizing their data capture to address this, but it’s a touchy area since this means you’ll have less information about how best to understand your transactions.
So, what do you see? Is the risk worth the data in the work you do?