A while ago, I did some data mining on my Netflix rental statistics. I tried to make those tools open so other people could study their own. I’m sad to say that those tools no longer work (Netflix changed what data was available and the format,) but I’m happy to say someone else has picked up the slack and made much neater tools.

Netflix provides RSS feeds for users to view their queues, recommendations, reviews, etc. Feedflix (feedflix.com) is a free service that mines those RSS feeds (you sign up with your private feed code,) to show you more information about how you use Netflix. It goes a step further to aggregate everyone’s queues and at-home movies to show data that Netflix doesn’t publish. (For example, National Treasure : Book of Secrets is currently at the homes of 55 Feedflix subscribers, is queued by 376 of them, and has been reviewed 8 times.)

Some of the data you can see without submitting your Netflix data. For example, I am in the 10% of surveyed users whose queue size is between 400 and 499. Almost 75% of the surveyed users return their movies in 9 days or less. Yet another way I feel like an outlier (at least I have nice hair.)

If you use Netflix, add in your RSS feeds so we can get a larger aggregate picture of Netflix users. (You’ll also appreciate knowing that you’ve had the 24 DVD for 46 days and each rental has cost you $12 a movie.) The site is maturing, and I’ve talked with the developer. He’s always looking for more data and more suggestions on how to manipulate it. If you liked my day of week calculations or cost calculations, those are in Feedflix now, so be sure to check it out.

I’m really impressed with the Feedflix data and how its improved my Netflix user experience. My only existing criticism is the lack of sample size anytime a percentage is mentioned. (For example, 23% of Feedflix subscribers return movies within 3-4 days… but 23% of what? 100 users? 10,000? A million?)