The current Campaigns & Elections TechBytes column went online today, and it’s chock-full of goodies for you fine people to enjoy. First up, Obama Chief Scientist Rayid Ghani and others talk a about the limits of data in political targeting. Remember that relentless data-driven campaign we’ve talked about again and again? It was far from perfect, at least as the staff describe it: “Ghani said that for every mention of ‘data integration’ on the campaign, he had ’20 caveats’ about how less-than-perfect it actually was in practice.”
The article explains why that was true, with reasons ranging from inherent limits in the data (“Ghani said that the Obama voter file was actually the smallest data set he’d worked with as a technology professional, in part because people vote so rarely.”) to the difficulties involved in building a system while you’re also using it (building an airplane in flight!). A resource allocation “better than random” was their goal…which despite the caveats, the campaign more than achieved.
Also in TechBytes: more detail on Obama’s data-driven TV ad-buying effort, which helped him get far more bang for the buck than his rival (though with one caveat: I’ve also heard it said that Romney’s ad-buying was particularly incompetent, which may skew the numbers). Watch for plenty of campaigns to try to use the same approach in the future, using data to target the voters they need to reach with a minimum of wasted ad impressions. For more, check out the column.