It’s about Little Data, not Big Data

Hyoun Park covers Business Intelligence for Nucleus and just published a research note on Little Data. I gave him a hard time a few months ago when I saw a coffee table-sized Big Data book on his desk. When books like that are being published there’s something wrong with the topic and Hyoun knew that. Analysts are skeptics at heart, and one of the challenges with the Big Data trend is what I’ll call the “Magic Pixie Dust” phase. In short, the pitch for Big Data is to span as much data as possible so you can (after sprinkling “Magic Pixie Dust”) find interesting relationships and make better decisions. That’s quite a leap from gathering lots of data to making a decision. As Hyoun correctly points out, the real challenge is narrowing the data to the fewest data points you need to make an accurate decision – the concept of “Little Data”. Put in a real-world context, when a guy walks into a frat party, he doesn’t really want to know the names, ages, and home addresses of every girl at the party, just the one most likely to talk to him and maybe a good opening line. Okay, I’ll work on the political correctness of my analogies. Little Data makes sense, and we’ll see a lot more from Hyoun on this topic.