The Netflix Data Machine

Jun 21, 2012 by Phil Simon in Data Management, Data Quality

I’ve written before on this site about Netflix’s gaffes. The TV and movie viewing site has made some interesting and unpopular decisions over the past year, although it seems to have righted the ship recently. Despite these missteps, there’s one thing that the company has done very well: manage its data, even if that meant holding contests to improve upon its own recommendation engine.

Consider the absolutely staggering amount of data that Netflix collects on its users. According to GigaOM, some of these statistics include:

• More than 25 million users
• About 30 million plays per day (and it tracks every time you rewind, fast forward and pause a movie)
• More than 2 billion hours of streaming video watched during the last three months of 2011 alone
• About 4 million ratings per day
• About 3 million searches per day
• Geo-location data
• Device information
• Time of day and week (it now can verify that users watch more TV shows during the week and more movies during the weekend)
• Metadata from third parties such as Nielsen
• Social media data from Facebook and Twitter

In a word, wow.

Possibilities Unlimited

Imagine the predictive possibilities of effectively harnessing this data. My industry friends tell me that, in the past, quite a few movie and TV “green light” decisions were made on the basis of some bigwig’s hunch. For instance, a VP at AMC or HBO opt to pick up or renew a show based upon gut feel, very early ratings, or preliminary focus group data. In fact, über-successful Seinfeld was effectively cancelled after six episodes until an NBC honcho went to bat for the poorly rated show about nothing.

The rest is history.

Now, let’s apply Big Data to the increasingly dispersed world of video content. Toss in funding platforms like Kickstarter to boot. It’s no exaggeration to claim that the very nature of TV show and movie development might fundamentally change.

We have already seen videos go viral. What’s to stop shorts of shows and movies from doing the same thing? What if a five minute sample of a show generates millions of views? Isn’t that a compelling reason for a studio exec to pick it up for five minutes? What if I know that viewers of Mad Men are 80 percent likely to give a show with Jon Hamm a try?

Simon Says

I’m not terribly bullish on the future of Netflix, especially since Amazon, Apple, and Google are getting into Netflix’s space. If Netflix goes down, though, it won’t be without a fight–and the company’s use of data might be its ultimate weapon.

Feedback

What say you?

Tags: Data Management, Data Quality, netflix