Netflix data reveals stark geographical contrasts in movie popularity

By Sarah Sluis

Netflix has long been public with their data, even offering a prize of a million dollars to the group that could improve its recommendation system. The contest, of course, required an open sharing of the data. The New York Times released a fascinating infographic last weekend that overlays Netflix rental patterns with Google Maps of twelve U.S. cities. I took it upon myself to analyze the results.

"The White Hole" For a number of reasons, the movies rented in a city center tended to be different than those rented in more suburban areas, leading many maps with a white hole in the center to indicate no "Top 50" rentals. Most obviously, Paul Blart: Mall Cop played well in areas outside cities, presumably populated with malls and Segway-using security guards. Nights in Rodanthe also played better outside the city, where older females might live. A lot of run-of-the-mill genre films and other drivel played well outside city centers (Dallas loved Knowing), while many art films played better in cities. Is it any wonder that Taken was a "surprise" hit, given that the movie is not in the Top 50 rentals for most of Manhattan and Brooklyn, the center of media? But exceptions prove the rule: Last Chance Harvey, an indie film, played best outside city centers. Perhaps the younger demographics in most cities bumped it out of the top 50 list? Children, too, can account for the "white hole" effect. Bedtime Stories played best outside of city centers. Pixar's Wall-E, however, had more rentals within the city. Here, I suspect a confounding factor. Families in the suburbs likely bought Wall-E (and rented the lesser-reviewed Bedtime Stories), while city dwellers rented Wall-E and passed by Bedtime Stories.

The "Made Locally" Effect Do you remember New in Town, the Renee Zellweger flop about a Miami transplant having a go of Minneapolis? Well, apparently the only people that do live in Minneapolis. The romantic comedy had no traction anywhere else. While most cities shunned romantic comedies--even well-reviewed ones, like The Proposal, did better further outside the city, Confessions of a Shopaholic was a modest exception. The frothy shopping film had a good ranking in the Upper East Side,

Midtown, and West Side neighborhoods in Manhattan, but was utterly absent in the East Village, Lower East Side, and West Brooklyn neighborhoods (like Williamsburg and Park Slope). Here, again, there are two possible solutions. The hipster-populated neighborhoods either saw the film in theatres, and that was enough, or they were simply too cool for the movie. Given the tiny box-office take, I suspect the latter. Other films appealed to niche populations tied to a geographical location, even if the movie wasn't set in that city. Appaloosa, a Western set in Albuquerque, did exceptionally well in parts of Dallas and L.A., but little business in the other big cities.

"The Black Boost" Films with black actors in lead roles did really, really well in the areas of cities with large black populations, while showing little or no "Top 50" activity anywhere else. Atlanta and Washington D.C. lit up for films like Lakeview Terrace, Traitor, Obsessed,

The Secret Life of Bees, Tyler Perry's The Family that Preys, Tyler Perry's Madea Goes to Jail, The Taking of Pelham 1 2 3, Not Easily Broken,

and Cadillac Records. That's a lot of movies. The South side of Chicago, areas of Dallas, spots all over New York City, and wee sections of Boston, Miami and L.A. also showed heavy activity on these movies. The exception was Seven Pounds, the Will Smith would-be awards movie that was dark red everywhere, indicating heavy rentals.

"The Blue State Boost" Surprise, surprise. People who saw Bush-lambasting W. were concentrated in blue state areas. Milk, set in San Francisco and appealing to liberal, gay-friendly, and awards film audiences, had high levels of rentals in the Bay Area and the liberal areas of other cities, though it bled more easily into other areas than W. Religulous did better in Western states, which have a reputation for being less church-going. Denver, followed by Seattle, was hot for the religion-mocking documentary.

After noting these trends, I'd also like to point out some confounding variables. People who subscribe to Netflix have something in common--they enjoy movies, have enough income to pay for monthly deliveries of said movies, and probably have some other demographic information in common, whether it's race, gender, age, income, or level of education. That can skew results. Netlfix is not the only sign of a movie's popularity in a particular location, since it doesn't know if someone saw the movie in theatres or bought the movie. As noted, that's why I suspected Wall-E was being rented in urban areas and bought (and thus off the map) in areas outside cities. That being said, this data can provide valuable information to distributors and exhibitors wondering why some films that do well at the box office flop for their theatres, and vice versa.