11 Points

Shazam Can Predict a Hit Song 33 Days in Advance (and the Future of Data-Driven Entertainment)
written by Sam Greenspan

Shazam is one of those apps that seems like magic the first time you use it.

For those unfamiliar, you fire up Shazam when you hear a song you like but don't recognize. You just press the button, it listens, and within a few seconds it gives you the artist and title. Really, it's like magic.



And like everything magical out there, we probably should've known it was collecting reams of data on all of us.

Can't fault them; I'd do it too. After all, as with all free apps, your data is as valuable as (perhaps more valuable than) any ad revenue it might generate. But this isn't a piece on privacy or data aggregation or "if you don't pay for a product, you are the product" -- I'm pretty sure those topics have been covered to death, had their corpses dug up and reanimated, then killed again. We've all tacitly conceded that we're fine with the tradeoff of sweet-ass free apps for giving up personal data.

This post is about the fascinating knowledge that can be gleaned from 100 million active human data points and what it means for the future of music.

Shazam's vice president of product, Cait O'Riordan, spoke at a conference last week and revealed a little bit about what Shazam is capable of:
  • Predicting a hit song 33 days before it starts moving up the charts.

  • Predicting which unknown artists are about to explode.

  • Finding which moment in a song is most likely to inspire someone to Shazam it.

  • Finding which artist on a collaboration resonated most with listeners. (For example, she cites Nicki Minaj on Monster getting more of a response than Kanye West or Jay-Z.)

Insane stuff. What could this mean for the future of music?

Short-term, it means record companies should use data from Shazam (and also data from services like Spotify and Pandora) to do everything from testing the performance of their songs to shaping their strategy around new artists. I would assume they already have some sort of contract deal where they do that.

Long-term, it could mean a much more seismic change: A change to music itself. Shazam already has enough data to (almost) definitively determine what makes a song popular; for what inspires people to say, "I like this song and I want to know more about it." The next step? Reverse engineering an algorithm for music itself. Coming up with a formula to engineer songs that are all but guaranteed to be hits. They already have the power to create that algorithm.

And it's not just music. Netflix knows more about our TV watching habits than the broadcast and cable networks ever did; their programming choices almost certainly contain some amount of statistics behind them. Amazon can link your purchases (and your ZIP code, age, and more) to what TV shows and movies you stream and what books you read.

Data-driven creativity is coming -- if it's not here already.

Every movie in theaters, every TV show on a network, every song from a record company, every book from a major publisher -- those will all be analytics-driven and sculpted. They have to be. Why take a chance on a bunch of executives' gut feelings when statistics can provide a 90 percent chance of creating a "hit"? You already see signs of this, like studios reducing the number of movies they produce to focus on superheroes, adaptations and remakes.

The goal for creators, then, is to find ways to work within that system; to marry their creative vision with the numbers for something both artistically and mathematically fulfilling. (It will happen. Not every time, but it will happen.) Data-driven creativity isn't, believe it or not, as inherently evil as it sounds. Good can come from it -- bigger budgets, better talent, more eyeballs -- it's just a new set of laws that require some fairly massive adjustments.

Meanwhile, for those outside the system, the Internet will still be here, providing a platform for the independent writers and directors and musicians, creating whatever they want in an unbridled, lower-stakes environment and letting the market decide its fate.

And sure, upon great success, they'll all be snapped up by the mainstream, but it's a good, fulfilling ride to get there.

tl;dr: How apps like Shazam will prevent a future Kazaam.


This post was originally published on Monday, May 11, 2015 at 11:00:00 AM under the category Web & Tech.

Facebookery?
11 Most Disappointing TV Shows of the 2000s

11 Movies You Didn't Know Were Based on Books

A Random Fact About a 90s One-Hit Wonder Is Now My Favorite Piece of Music Trivia Ever

11 Most Unpopular Emoji

11 Points About Our I Hate Britain Video

Archive of all Web & Tech posts
A Ban on Everclear Grain Alcohol Is Bad News... For the Violin Industry?
A Ban on Everclear Grain Alcohol Is Bad News... For the Violin Industry?
Published Thursday, July 6, 2017 at 11:00:00 AM under the category Food & Drink
11 Pairs of Classic Movies That Were Surprisingly Released on the Same Day ('90-'94)
11 Pairs of Classic Movies That Were Surprisingly Released on the Same Day ('90-'94)
Published Thursday, June 22, 2017 at 11:00:00 AM under the category Movies
11 One-Hit Wonders Whose One Hit Was a Cover
11 One-Hit Wonders Whose One Hit Was a Cover
Published Friday, June 9, 2017 at 11:00:00 AM under the category Music
11 Nintendo Video Game Systems, Ranked By How Many Games They Got
11 Nintendo Video Game Systems, Ranked By How Many Games They Got
Published Friday, June 2, 2017 at 11:00:00 AM under the category Games
11 Songs That Just List Off a Whole Bunch of Cities
11 Songs That Just List Off a Whole Bunch of Cities
Published Friday, May 19, 2017 at 11:00:00 AM under the category Music
11 Coincidental Wardrobe Choices That are Evidence of Cosmic Intervention
11 Coincidental Wardrobe Choices That are Evidence of Cosmic Intervention
Published Wednesday, May 10, 2017 at 11:00:00 AM under the category Misc
Full Archive
Follow
11 Points

Mailing list


Twitter


Facebook


YouTube


Pinterest


RSS