The Structure of #BIGdata

Enjoy a new 6 minute introduction to #BIGdata by @JRMigs

"Back when I started, not only didn't we have #BIGdata, we had to convince our clients that they might want to save more than basic address information. Lets start piling it up and we'll figure out what to do with it later. If you, knowing all you know about you, can barely predict what YOU will do... how can I, as a predictive modeler, knowing almost nothing about you, expect to predict you? What is the difference between HARDdata and SOFTdata?

10

10

Why don't Marketers decide better?

Marketers Flunk the Big Data Test - Patrick Spenner and Anna Bird - Harvard Business Review
Great article, especially the part about a team... keeping bright shiny objects out of view... :). The truth is that good use of data does require more than one type of brain. We have very very careful people in charge of loading and processing data, clever systems people who write very very fast procedures to generate hundreds of variables and oh, there is that weird wizard-type guy who we rarely take out in the daylight (me). Data is not self-revealing, even with validation, noise and correlation can be confused - and causality is never really in there. As Dr. Ed Lemer, UCLA, http://bit.ly/O6RXNh said to me, "You don't violate statistical principles, you have a higher standard... that it makes sense." And almost none of the computers I have know, could tell me what would sell. Until we get that built into our use of #BigData, there will still be that strange alchemy of science and magic that makes marketing the fun that it is. bit.ly/McLWsp @JRMigs 

Another interesting related article. "True direct marketing offers the greatest opportunity to attain marketing certainty, but only if you test. Until we reclaim that message, we’re just another piece of the ad mix. And without testing, we’re not nearly the most important piece."