Streaming Music Grows Up

AUSTIN (TheStreet) -- Almost since its inception, the streaming music space has been rife with disagreements over technology, over man vs. machine approaches and over subscription vs. free models.

But in sign of a maturing sector, SXSW presentations and interviews this week -- with representatives of Beats MusicSpotify, Echo Nest and Pandora  (P) -- have shown those differences are melting away, forming a single set of problems and solutions that all the players are addressing.  In panel discussions on issues like the necessity of including a human component in the presentation of music to an audience, or the way streaming music is changing local music markets into one international bazaar, the big problems facing streaming music companies sounded remarkably the same from one company to the next.

In particular, the future of the sector will involve a greater sophistication of existing algorithms, a trend toward putting a human face on the music presentation process and sweeping data collection to better target individuals on an increasingly local level but on an increasingly international scale.

In a panel provocatively titled, "Man Vs. Machine: The Curation Dilemma," panelists seemed to agree both about the "dilemma" and its solution: humans need humans at the controls, helping guide and refine the music selection process for listeners in various ways, both in the deep background data collection and on the front end, as celebrities making recommendations for instance. Ian Rogers, CEO of Beats Music, referred to this as "curation by trusted sources."

When faced with the option of what to listen to, users can choose from any of millions or hundreds of millions of songs, depending on the service. Using algorithms is important to help them make sense of their choices. But if the recommendation is purely mechanistic, it's less helpful than if it comes from a specific person.

As Rogers put it in an separate interview, "As human beings we actually really value the cultural currency of music, who recommended you something -- it's impossible to separate the recommendation from who recommended it. Music is never without cultural context."

"I don't think an algorithm can ever give you that," he added.

While Beats Music is a subscription-only service, Rogers also wasn't ruling out a free, ad-driven component at some point in the future.

Beats launched as a spinoff from Dr. Dre's Beats Electronics only six weeks ago with much fanfare. The company builds its service partly on user data and song data it inherited from Mog, the music service purchased by Beats in 2012 that served as the foundation for Beats Music. But what Rogers and his company choose to emphasize -- both in the user interface and in its marketing -- is the human component, the celebrities and music personalities choosing the playlists and making recommendations.

At Pandora, the machine recommendations carry the stamp of a deeper human involvement. Humans examine data and "groom the collection of music" for each listener, according to Chief Scientist Eric Bieschke, speaking at the panel discussion. But that doesn't necessarily preclude the celebrity playlists touted by Beats Music or offered at Spotify, as he told me later in an interview.

"You can't ask a machine why this particular song made you cry," Bieschke said. "There's a place for [celebrity or user playlists]. I don't think it's going away but en masse it's not going to work."

Without more sophisticated algorithms that can efficiently and effectively identify a range of users' tastes at a particular moment or for a particular purpose, playlists can quickly become as cumbersome to deal with as the songs themselves.

Those algorithms, in turn, need to have more accurate information, collected from mobile devices and through tracking usage. Where you are, what the weather is like, the time of day -- all of that can help the machine identify the context of a user's music listening and better enable it to make recommendations.

The goal, as Bieschke put it, is "to connect you and only you with the thing you like to hear."

Scaling the delivery of music to millions of users and yet having the music experience be a personal one is the single greatest ongoing challenge facing the sector and it is multifaceted, with pieces in data collection, pieces in a detailed, musicological analysis of music and pieces in ethnographic and cultural understanding of particular areas.

As an example, Echo Nest CEO Jim Lucchese noted in a private interview that the channels through which songs become popular, the context through which they emerge, can vary in ways that affect streaming selection.

"Things like the fact that people discover popular music in India though film in way that's unique to India, creates an entire new data structure challenge," he said.

At the moment, the industry is relying on western-centered cultural assumptions, but that has to change as an ongoing explosion of digital popular music in Latin American, Asian and African markets is promising to dramatically shift the music industry as a whole. The streaming sector is a pivotal player in that shift: as handheld devices become readily available, some societies with limited Internet access through landlines are skipping past the downloable-digital-file stages experienced in the U.S.'s technological development, leap-frogging directly into streaming over mobile.

While addressing that growing need represents a goal for most of the streaming services, it is a more immediate need for Spotify, Echo Nest's new parent company and Lucchese's new bosses. According to Ken Parks, Chief Content Officer, Spotify is already active in 55 countries around the world and is expanding into many others.

Parks told TheStreet in a private interview that in addition to some packaging changes to the user interface coming in the short term, the company is focused on reaching directly into markets where the challenges to a Western company are greatest. For that, they need tools to analyze and present effectively in the local cultural context.

Even with their international advantage over the rest of the streaming players, the challenge remains a huge one for Spotify.

"We've got to be best in class at every instance," Parks said.

Partly those challenges will be met by parsing the flood of personal information that is becoming available on each consumer on the planet over the coming years. That "explosion of data" as Lucchese termed it, will be useless without the tools to parse and apply it in ever more intelligent ways.

The streaming sector's challenges, goals and even methods of operation are becoming more homogenous across the board, unifying them on a deeper level in the challenge of shaping the global music business landscape for the rest of the century.

As the purchase of Echo Nest shows, the sector is ripe for consolidation, but brand identity -- once obvious -- is going to become increasingly more imperative. As their functions, philosophies and technologies become more similar, a growing challenge is emerging to create tools and marketing that will more clearly differentiate one service from another for the end user.

-- Written by Carlton Wilkinson in Austin, Texas

If you liked this article you might like

Wendy's, Roku, GM, AT&T and 3M - 5 Things You Must Know

Wendy's, Roku, GM, AT&T and 3M - 5 Things You Must Know

Internet Stocks' Momentum Won't Stop Anytime Soon: Goldman Sachs

Internet Stocks' Momentum Won't Stop Anytime Soon: Goldman Sachs

A Fujifilm Deal Could Derail Carl Icahn's Xerox Overhaul Effort

A Fujifilm Deal Could Derail Carl Icahn's Xerox Overhaul Effort

Spotify Business Model Spells Challenges for IPO and M&A

Spotify Business Model Spells Challenges for IPO and M&A

FCC's Dismantling of Net Neutrality Could Spur New Wave of Telco M&A

FCC's Dismantling of Net Neutrality Could Spur New Wave of Telco M&A