While reading this interesting article about Dstillery, a big data intelligence company that “demystifies consumers’ online footprints“ to find the perfectly fitting ad, I had, yet again, to think about how fascinating data analytics can be.
After only a few years with data analytics, we have Health Analytics, HR Analytics and know data is also influencing politics more and more. When the caucus in Iowa took place, Dstillery did an amazing job in analyzing the different voter types.
The company that I mentioned above gets information from people’s electronic devices via ad networks.
What you might not know: when you open an application or search for something on the internet, something happens without you even noticing it. There’s a tremendously fast auction that happens, where different advertisers bid to get to show you an ad on your own device.
The highest bidder is the one who thinks that you are the most valuable. To calculate this, your device sends them key informations about you, sometimes also with an identifying code and the location.
So what happened in Iowa was the following: Dstillery looked at all the different auctions that took place on electronical devices near the caucus locations.
Nearly 16,000 devices were spotted this night, thanks to people who granted permission on their location to those apps etc. that served ads. The system saved the identifying codes and searched for characteristics that could have been associated with IDs in the different caucus locations (Republican/Democrat).
This is definitely an amazing way to combine online and offline tracking.