12 January 2010

a free idea for someone much, much smarter than me

Over at The New York Times (maybe you've heard of them?), there's a super-cool interactive map of Netflix rental patterns in a couple of major cities. Go ahead. Check it out. I'll wait.

There's all sort of interesting stuff there. You can, for example, see that a film like Rachel Getting Married does really well in some parts of NYC (like Brooklyn) and isn't really rented at all in other areas (like the Bronx). And while your first reaction would be that Yankees fan are idiots (and you'd be right), there's more to it than that.

This is a lot of data. Tons of data, actually, and that's clearly just a fraction of it.

Now, for those of you who aren't familiar with the work of Nate Silver, go read this summary of his processes for figuring out polling data and elections. It can be fairly heady stuff, but this is the important part:
In reality, there is no such thing as national polling movement. Instead, you have millions of individual voters making up their minds in 50 individual states and the District of Columbia.

Like-minded voters, however, can be presumed to change their candidate preferences in similar ways. For instance, relative to national trends, election results in Massachusetts have historically been 90 percent correlated with election results in Rhode Island.

Our simulation accounts for this tendency by applying a similarity matrix, which evaluates the demographic relationships between different states by of a nearest-neighbor analysis as described here. Our process recognizes, for instance, that as the polling in Ohio moves, the polling in a similar state like Michigan is liable to move in the same direction. On the other hand, there may be little relationship between the polling movement in Ohio and that in a dissimilar state like New Mexico.

Do you see where this is going yet?

Let's say you had a film that was accurately described as Rachel Getting Married meets The Wrestler that you were promoting in NYC. You could sort of eyeball where in NYC you'd have better luck promoting it and where you'd be wasting your time. And they're pretty similar (and pretty drastic), so it wouldn't be too hard. You'd probably start putting your resources into Brooklyn, right? But this is only a few films in a few cities. Imagine if you had access to the entire database for the whole country. You might discover that our mythical film would play really well in Birmingham but not in Atlanta.

It's long been a hallmark of Hollywood that if you like a film, there's a pretty good chance you'd like a similar film. It's human nature.

So imagine this: you have a film you're looking to self-distribute. You go to a website that specializes in the Netflix map database and you punch in your closest match films, a minimum of 2 but up to 5 (5 would likely be more accurate) and it'll spit out a nationwide map just like one on the NY Times webpage. You can bring it down to a zip code level. How valuable would that be to an indie film? You'd know exactly where to target your film for the best results. Think of how much time and energy you'd save.

It'd be difficult, sure, but it'd be super valuable. And it wouldn't just help indie films. Theaters could use that data to better program films (even multiplexes!), as could festivals. It's a mountain of information that no one is using (except for Netflix).