Last fall, The New York Times resident numbers geek Nate Silver provided a lesson in predictive analytics for the whole world to see – crunching big data to predict, with almost pinpoint accuracy – the winner of the U.S. presidential election.
The success of this high-profile project thrust big data analytics into the limelight, but there are many, somewhat more mundane applications, but with even more unforeseen revelations. Such applications are explored by Eric Siegel in his latest work: Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die. Siegel points out some wild, unexpected revelations that have come out of predictive data analysis. Interestingly, the explanations for these unexpected outcomes seem to amount to educated guesses, which opens up the next great unexplored frontier for data scientists and analysts – we’re learning ‘what’ takes place, but we still need to understand ‘why.’
Siegel is well aware of this, observing that the explanations offered “do not explain a single thing. Correlation does not imply causation.” He adds that “when applying predictive analytics, we usually don’t know about causation, and we often don’t necessarily care. For many predictive analytics projects, the objective is more to predict than it is to understand the world and figure out what makes it tick… Predictive analytics operates with extreme solution-oriented intent.”
So with this in mind, here are some odd-ball results organizations have seen from their predictive analysis, and possible – emphasis on possible — explanations:
Vegetarians miss fewer flights: A major airline discovered that “customers who pre-order a vegetarian meal are more likely to make their flight.” A possible explanation is that the special meal order “establishes a sense of commitment.”
Shopping habits predict reliability as a debtor: “If you use your credit card at a drinking establishment, you’re a greater risk to miss credit card payments.” Use a credit card at the dentist, lower risk. Use a credit card to buy felt pads that affix to chair legs to protect the floor, lower risk.” This research by a financial services company suggests “more cautionary activity such as the dentist reflects a more conservative or well-planned lifestyle.”
People consistently considered attractive get less attention: An online dating site found that “online daters rated with a higher variance of attractiveness ratings receive more messages than others with the same average rating but less variance.” The explanation is “people often don’t feel they have a chance with someone who appears universally attractive. When less competition is expected, there is more incentive to initiate contact.”
A job promotion can lead to quitting. A major IT vendor found that “promotions increase the risk an employee will leave, unless accompanies by sufficient increases in compensation. Promotions without raises hurt more than help.” The bottom line: “increased responsibilities are perceived as burdensome if not financially rewarded.”
Mac users book more expensive hotels. “Orbitz users on an Apple Mac spend up to 30 percent more than Windows users when booking a hotel reservation.” The explanation offered is that “Macs are often more expensive than Windows computers, and Mac users may have greater financial resources.”
Banner ads affect you more than you think. “People who see a merchant’s banner ad are 61 percent more likely to subsequently perform a related search, and thus drives a 249-percent increase in clicks on the merchant’s paid textual ads in the search results.” Blame it on the subconscious effect of advertising, Siegel relates.
Sometimes, more web visits mean a potential loss in business: More frequent visitors to wireless sites are more likely to defect when their contract is up, a major telecom found. By way of explanation, those more likely to jump ship are checking on early termination terms and fees.