Our Digital Tracks
The originally French word routine comes from the word route (meaning road). Literally, routine means the road normally taken, and, figuratively, it is the habit of doing something in the same way. So, both literally and figuratively, a routine is a repetitive behavior in our lives. This repetition — our habits — creates a structure within which we feel security and familiarity and can become automatic in our minds within just a few weeks. In 2006, researchers at Duke University reported that approximately 45% of our daily behavior is a repetition of some type.
In an age where we leave digital tracks wherever we go — which websites we visit; which e-commerce products we are interested in; which sports, movies, and shows we watch; what music style we listen to; which roads we travel on — more than ever, our habits can be watched, quantified, and measured. Smart cities, for example, can make use of traffic information to plan new roads, reverse the direction of certain streets at specific times, and reduce bottlenecks, thus decreasing pollution. Your Internet browsing sessions can be interrupted by ads geared specifically to you — for example, an ad for the world parachuting championships, based on a past purchase or a trip connected to parachuting locations you may have made.
The processing power that is now available enables corporations to detect subtle changes in consumers’ habits — and this has become a significant business opportunity: A graduation, change of city, marriage, pregnancy, or divorce are all indicators of possible changes in consumption habits. A story that reflects this new dynamic is told by American journalist and writer Charles Duhigg in his 2012 book The Power of Habit.
In 2018, the US-based retailer Target had nearly 2,000 stores nationwide. Attuned to the habits of its customers and seeking new opportunities to build loyalty, the company’s analysts were tasked with detecting when a woman became pregnant, using data available in their systems. Pregnancy significantly modifies consumption habits, so the earlier Target learns of this event, the faster it can act with relevant offers.
The team of technicians analyzed historical purchasing patterns of women who signed up for the baby registry on the company’s website, letting big data techniques detect correlations that would show products with a likelihood of indicating a pregnancy. Among the 20 products the system yielded were moisturizing creams and dietary supplements. Based on the dates of these purchases, Target identified not only their pregnant customers, but also their stage of pregnancy. The next step was to introduce a program that offered products recommended specifically for each trimester.
But that wasn’t the end of the story told by Duhigg. He recounts the story of a father who went into one of the outlets demanding to speak with the manager, furious that the retail chain was sending his daughter (who was still in high school) coupons for baby clothes and cribs. The angry father asked the company to stop, saying it would encourage his daughter to get pregnant. The issue was escalated inside the company, and a few days later the father was contacted by a representative. During their conversation, the father ended up telling the agent that after his visit to the store, he had learned that his daughter actually was pregnant, and that the offers were therefore appropriate.
Next time, we will continue to discuss Big Data, diving into its predictive power. See you then.
