Facebook status is a great tool for holding a conversation when you’re doing some lightning-fast user research. What other ready-made tools do you use to source insights from your crowd?
Archives
Categories
Tags
-
RSS Links
Meta
NPR has a great podcast on what credit card companies are doing to try and predict whether you’ll be a bad credit risk. Credit card companies are basically taking your purchase history and data mining it (in a similar way to Netflix), to profile you. One study cited is Canadian Tire (a Canadian retailer, kind of like Home Depot), who took all the purchase information from their credit cards and ran an analysis on it to find correlations between purchase and credit risk.
The results show two extremes. People who purchased the product premium wild bird seed are most likely to pay their bills on-time. While people who buy chrome-skulled accessories for their car are the worst credit risk, defaulting on their bills 4 times in a year on average.
Whereas before the recession, this information may have been used to market more credit cards to you, nowadays credit card companies are scrambling to get their money back from all those chrome-skull afficionados.
Since the 1980s, credit cards have been making the bulk of their money from bad money management having realized that “the biggest profits didn’t come from people who always paid off their bills but rather from less-responsible clients who never paid their entire balance, and thus could be milked through silently skyrocketing interest rates, late fees and other penalties.”
It turns out, after the recession hit, people couldn’t even afford to pay those interim late fees anymore.
This leaves credit card companies fire-fighting to recover all that debt, and one approach that’s working is to establish an emotional connection with customers. Call center staff are becoming agony aunts to debtors, offering words of comfort and advice, trying to solve problems rather than just demand their money back.
“Today the goal is for customers to get a warm-and-fuzzy feeling from their credit-card company,” said Carl Pascarella, a former chief executive of Visa USA. “If we have a deep relationship with you over a range of products and experiences, if we trust each other, you’ll listen when we give you advice.”
Hopefully beyond the recession, card companies will adopt this design approach as a way to prevent untenable debt, rather than just dealing with its consequences.
For example, how might my credit card company offer tools to help me manage debt, or become a partner in my finances rather than a pain point?
Read Charles Duhigg’s New York Times article for more details
After reading this New York Times article on the Netflix Prize , I thought it might be cool to register and play with some movie data.
Netflix is offering a chunk of their movie rating database (2 GB worth) for download, it contains ratings for 17,770 movies and TV shows from about 300,000 customers. With it, contestants in the Netflix competition have to write an algorithm that can predict how customers will rate movies they haven’t seen before, thus enabling Netflix to better recommend movies. This algorithm has to do 10% better than Netflix’ existing recommendation engine. And the prize is a million dollars.
And people are close. Really close. If you check out the leaderboard , the top team PragmaticTheory is at 9.65%!
So given that my hard drive only has 2.5 GB free (after the data download from Netflix), it was pretty obvious I didn’t have the processing nor storage power to handle all the data at once. I decided to start with a very very small subset (like 11 movies) and get a little taster for what it would take:
Pirates of the Caribbean: The Curse of the Black Pearl, Rushmore, Miss Congeniality, Pretty Woman, Forrest Gump, Twister, The Patriot, Independence Day, The Day After Tomorrow, Con Air, The Green Mile
The process was to take each person on file, look at how they had rated all these movies and try to classify them into a group. Completely unexpectedly, the program created a very simple model that revolved around… Forrest Gump.
Forrest Gump turns out to be a strong predictor of whether people will like other movies (ie. people who love Forrest Gump also love the Green Mile, and surprisingly also really like The Patriot). But it doesn’t work in reverse, other movies are not as good a predictor of whether someone will like Forrest Gump. I wonder if there are a handful of movies which act as strong predictors, from which preferences for other movies can be surmised. Forrest Gump is effectively at the top of a giant movie decision tree, the thickest of trunks leading to more subtle differences in movie tastes.
A good analogy is the 20 questions game. Where the power of a decision tree can narrow down to an exact thing, through asking questions that would make the biggest differences first.
And decision trees are set to become a hot topic again with the creators of flickr soon to launch a new web service all about decision trees, called hunch.com. Only they’re using humans instead of computers to build’em. Hunch lists any decision you could make (including ‘What movies should I watch?’), building giant decision trees based on user contributed questions. Then (I’m guessing) it uses statistical analysis to figure out what questions sit at the root, the trunk, the branches.
