Predictive Analytics - Chapter 6

Use Cases in Other Contexts

In Comparison: Use Cases of Predictive Analytics in Other Contexts

Predictive analytics has fueled efficiencies and innovations in a number of different fields. It lets organizations better answer the questions that are most important to them. Here are a few examples:

What sicknesses are you most at risk for?

By looking at a patient’s demographic background, economic situation, and medical history, medical professionals are better able to predict what sorts of maladies someone may be at risk for in the short, medium, and long term. Patients can then alter their diet and lifestyle accordingly to combat problems before they become an issue.

Is this person likely to pay their debts?

Credit scoring is one of the longest standing uses of predictive analytics. Credit scores such as FICO and other ratings gauge the creditworthiness of an individual and evaluate whether lending that person money is a sound investment. Having this sort of predictive capability allows banks to lend money to a larger number of people with greater confidence as they can set interest rates according to the individual, rather than a blanket rate for all customers.

Is that person likely to vote?

Political campaigns have previously relied on charting and tracking historic election results and voter turnout. Predictive analysis allows campaigns to focus more directly on an individual voter’s preferences and likelihood to go to the polls, and then segment similar voters into groups. Resources are scarce on the campaign, so candidates can now spend their money with greater efficacy only on those who are both likely to vote, and are most likely to support them. These predictive models of support and turnout can also be updated throughout the campaign to incorporate new polling numbers and the results of voter contact points (such as door-to-door canvassing, phone banks, and online interactions).

Which movies do you like?

Netflix and other content providers have realized that understanding the preferences of their users is central to their entire business, and the better job they can do of predicting what someone is likely to watch next, the better they can present that content to the customer. These recommendations drive user engagement and inform company strategy. Netflix even offered a $1 Million prize to anyone in the world who could improve their “movies you might like” prediction engine by 10%.