1. What’s the real value of big data?:
The Obama administration’s report last week on the need to consider restricting how Google, Facebook and other Internet powerhouses collect and use big data reminds me of a story I’ve been hoping to see for a while: How much does this collecting and slicing and dicing of big data actually help advertisers and marketers?
I get the idea that a woman who lives in New Jersey and has accessed information online about baby carriages makes a great target for advertisers selling other baby or maternity products. But do marketers really benefit from data that they buy that goes way beyond that — that zeroes in on what other websites she has been to, where she buys what online, where a location service says she has physically been lately or whether her Gmails refer to different products or subjects?
Two years ago, I was in an audience of media and marketing people mesmerized by a presentation from a Yahoo data expert who promised that his firm could target, to take one example, “men who had shopped online for a BMW and also been to a New York Giants football game in the last year.”
Each of those attributes might be worth something. Does putting them together really matter?
Maybe. But I’ve never seen a story that delves into the costs of using all that multi-grain data compared to the results it produces. How much more expensive are different levels of targeting compared to the results they actually produce? When does the big-data-based super-targeting become not only a privacy issue but also an expensive, alluring technology solution in search of a problem?