Automated Recommendation: Last.fm, Pandora and The Echo Nest
The dominant approach is to use the power of data and algorithms to understand the relationships between songs and listeners. Last.fm, Pandora and Echo Nest are the most established players in this field.
It has been awhile since Last.fm grabbed headlines the way Spotify and newer upstarts do today, but the service still offers one of the best music discovery tools out there. It monitors an individual's digital listening habits - from mobile devices to desktop players, even tracks streamed from the browser - and compares them to the demonstrated preferences of other listeners. The CBS-owned service provides an open API that developers have used to build all kinds of apps and mashups. It has remained relevant by weaving in as many other products and services as possible. Last year's Spotify integration, which brought the Last.fm user interface and functionality directly into Spotify's desktop client, is a perfect example.
Longtime Last.fm competitor Pandora is another veteran that still deserves recognition, even amid competition from newer entrants such as Slacker Radio and Songza. Pandora's Internet radio service uses a complex, partially human-informed algorithm that relates songs to one another. Its Music Genome Project goes deeper than simple artist-to-artist matches based on listening habits. It drills down to specific musical characteristics, from tempo to whether or not the song has an electric piano solo. This granular, detail-by-detail matching of songs makes for one of the most effective semi-automated ways to discover music digitally. Whereas Last.fm is great at surfacing unfamiliar artists, Pandora does a better job of suggesting unfamiliar songs.
Earlier this week, Spotify took aim at Pandora by launching its own free Internet radio service. Unlike Spotify's core streaming service, the new offering is available to nonpremium users via their devices. Spotify Radio is powered by The Echo Nest, which fuels dozens of music apps, including Clear Channel's iHeartRadio. With over five billion data points about artists and songs, The Echo Nest analyzes more data than Pandora, and it applies more automation involving data mining, acoustic analysis and machine learning. The resulting track-to-track transitions can be less sonically consistent than what Pandora comes up with, but it rivals Last.fm when it comes to artist discovery.
Social Recommendation: Friends Know Better Than Machines
As effective as these automated suggestions can be, it's still hard to compete with recommendations from people you actually know. That's why many music services are plugging into Facebook's social graph and integrating with Twitter.
This area of music discovery is still maturing, however. Last year's deep "frictionless" Facebook integrations with Spotify, Rdio and other streaming services created more noise than new favorites. Spotify's log, posted to Facebook, of individual listening behavior revealed what friends were listening to, but it didn't hint at why they should care. Meanwhile, the "group listening" trend that exploded last summer let users congregate in virtual rooms and take turns DJing tracks for one another. Represented by Turntable.fm, this category may not have lived up to its initial hype, but there's still plenty going on in this space. Soundrop, another group listening app, is built directly into Spotify.
Songbird uses data from Facebook likes and connections to pull appropriate audio from sources like YouTube and SoundCloud. The company aims to capitalize on what it perceives as the limitations of services like Pandora by relying more heavily on the social graph. Songbird does a really good job of pulling relevant artists from a Facebook profile, but it's not great at presenting new content. The service has some growing to do, but the hype surrounding it isn't unwarranted.
Not to be outdone by its social media elders, Pinterest is getting into the music discovery game as well. While the site is better known for sharing photos and other images, it's also used to post music from sources like YouTube and SoundCloud. It's not yet a major driver of music discovery, but considering its rapid growth, artists experimenting with Pinterest probably won't regret it.
Tastemakers and Human-Fueled Curation
The automated, social media-fueled approaches have their strengths, but they're by no means perfect. Computers alone are not likely to crack the music discovery nut anytime soon. As it turns out, we still need human brains to do some of the listening and interpretation that goes into music recommendations. That's why Pandora's algorithm works so well and why Songza has such promise. Both services, to varying degrees, rely on human intervention to help curate and cross-reference songs.
This need for human insight is why music blogs remain one of the most popular ways for people to discover new albums. The founders of Shuffler.fm are well aware of this and built their music-aggregating "audio magazine" to pull tracks from a wide selection of influential blogs. The result is a huge collection of genre-oriented channels featuring new and popular songs, curated by a virtual panel of Internet tastemakers.
When Spotify launched its third-party app platform late last year, it included services like Last.fm and MoodAgent, which analyzes and recommends music based on the mood it conveys. But it also launched with editorial partners like Rolling Stone, Pitchfork and We Are Hunted, emphasizing the important role that human critics still play.
Algorithms and APIs can do amazing things. But at the end of the day, determining what music - as well as art, movies, books and other media - people will like still requires human beings with real ears connected to real brains. The future of music discovery will rely on both.
0 komentar:
Posting Komentar