Spotify’s Got Audience Segmenting Chops, But Ad Targeting Still To Come

spotifyWhile music streaming giant Spotify has the capabilities and the insights to offer marketers granular audience segmenting and hyperlocal targeting, its focus on user experience supersedes the development of advertising products, said Gary Liu, head of Spotify Labs.

The user experience is “definitely a primary focus if not the primary focus for our ad platform product,” Liu said. “Our goal has always been to participate in the consumer’s music discovery experience.”

While bombarding listeners with ads might compromise that experience, Liu added, Spotify is still exploring how it can help brands interact with its listeners.

Still, the company’s stance on advertising hasn’t changed significantly from when AdExchanger spoke to its former Europe VP, Chris Maples, a year ago. “Our philosophy is that we will deliver our products and grow our business first and then scale our ad products as well,” Maples said then.

The roughly 75% of Spotify’s clientele that use its free service encounter audio ads every five or six songs, or three minutes of advertising for every hour of listening. Beyond that, advertisers can purchase banner ads or homepage takeovers.

For brand-specific advertising, Spotify can compile a curated playlist (even paying users who get an otherwise ad-free experience are exposed to this) or advertisers can build a microsite that lives inside the Spotify user’s dashboard. Only audio ads are available across all platforms, which include mobile, tablet and desktop versions of Spotify.

Spotify also has tracking capabilities, though it has yet to launch massive brand campaigns based on these insights.

“We track user behavior throughout the day,” explained Liu. “If users are listening to, say, electronic dance music every morning, early in the morning, Spotify can be pretty certain that the user is running or exercising, and depending on the frequency and repetition of listening habits, we can know precisely the best time to serve an ad related to fitness.”

Spotify calls these segments “musical attributes,” logging a user’s evolving musical preferences throughout the day, taking into account considerations like mainstream vs. obscure.

“When you think about what you see from a lot of creative researchers for advertisers, what they’re trying to get at is lifestyle attributes. Music is a great representation of that,” Liu explained.

The next, and much larger, piece of Spotify’s audience segmenting capabilities is tied to its five-month-old acquisition of The Echo Nest, a music intelligence platform that previously worked with a number of large digital music services (including Spotify), archiving data on music and music fans.

“All of the information we have about the content allows us to glean really deep insights about the consumer of the content,” said Jim Lucchese, CEO of The Echo Nest. “Our business has been to take all of this information, put it in an app development platform in order to power hundreds of music apps and figure out how we can work with Spotify and brands to help enable these experiences.”

This is where Taste Profiles come in, a tool designed to help brands target key audiences based primarily on listening activity. Taste Profiles take into account what listeners are sharing, favoriting or adding to playlists, in addition to matching sonic attributes to the time of day.

The technology is twofold, explained Lucchese. The first piece analyzes psychoacoustics, or how the construction of a song relates to perception and physiological effects, by tracking the pitch, tempo and time signature of a song.

This information helps Spotify predict a user’s identity and lifestyle. One potential data goldmine could be the login information Spotify gleans through Facebook. But the company declined to comment on how exactly it’s able to leverage that data.

“The advertising industry exists around music because we know intuitively that music is a great proxy and predictor of a user’s lifestyle,” added Lucchese. Given a user’s musical preference, Spotify execs say they’re able to predict likely affinities to being a gamer, an athlete or a parent.

“We’ve built predictive models to determine how particular lifestyle attributes can be useful to brands. By correlating all this data, we’re able to draw predictions. We can come in and give the advertising partner a lot more insights,” said Lucchese.

The Echo Nest overlays this with cultural analysis, which it gets by tracking and analyzing web conversations via natural language processing.

These elements help Spotify figure out what will likely become popularized. Taste Profiles help Spotify pinpoint users who listen to artists before they’re widely know, which can be useful for marketing forecasts.

“In the advertising context, specifically around targeting, these are two ways that we’ll be applying Taste Profiles in the future,” explained Lucchese, estimating a 2015 timeline to applying audience-level insights to Spotify’s ad products. “We have already begun working with some of our closest clients on the most impactful ways to deploy our capabilities in the ad stack.”

With Google’s forthcoming music service, YouTube Music Key (which boasts some eerily familiar offerings to Spotify’s), the music streaming space is about to get a bit more crowded, which could create greater competition for brand dollars, leading to improved advertising products.

Enjoying this content?

Sign up to be an AdExchanger Member today and get unlimited access to articles like this, plus proprietary data and research, conference discounts, on-demand access to event content, and more!

Join Today!