Get fresh updates from Hortonworks by email

Once a month, receive latest insights, trends, analytics information and knowledge of Big Data.


Sign up for the Developers Newsletter

Once a month, receive latest insights, trends, analytics information and knowledge of Big Data.


Get Started


Ready to Get Started?

Download sandbox

How can we help you?

* I understand I can unsubscribe at any time. I also acknowledge the additional information found in Hortonworks Privacy Policy.
closeClose button
prev slide
Startups Can Leverage Big Data for Big Results
February 07, 2018
How Using Big Data in Manufacturing Provides a Competitive Advantage
Next slide

Driving Customer Engagement Through Big Data and Music Discovery

Pandora. Spotify. Apple Music. Tidal. These are a few of the biggest names in streaming music. Livestreaming has forever changed how we listen to music, and the disruption these platforms are spreading through the music industry is only just starting. The next area that’s ripe for sweeping change is the use of big data and music personalization for live-stream listeners.

How are streaming services using data to understand what their customers want to hear? How are they using a wide array of data sources to make better listener recommendations? What insights do their personalization efforts provide other consumer-focused businesses?

Livestreaming “On a Scale That We’ve Never Seen”

Live-streaming music is the platform consumers demand, and it’s rapidly displacing physical sales (CDs) and downloads (via services such as iTunes). The Nielsen Music U.S. Mid-Year Report noted only halfway through 2017 that on-demand audio streams reached over 184 billion streams, a 62.4-percent increase over the same time period in 2016. David Bakula, the senior vice president of music industry insights for Nielsen, noted, “The rapid adoption of streaming platforms by consumers has generated engagement with music on a scale that we’ve never seen before.”

The key phrase to note is “engagement with music.” Subscription-based streaming has made discovery of new music easier. For lesser-known artists, it’s easier to find an audience and less costly to produce and distribute music. For consumers, it allows them to listen to as much music as they like for as long as they want. Their options aren’t limited to a download. They can listen to anything, anywhere.

Distributors have stopped selling ownership, and instead sell access. The barriers to engagement have dropped between musicians and listeners. How are streaming services using big data to improve customer engagement with their services?

Spotify’s Approach to the Personalization of Streaming

There are different approaches to combining big data and music personalization. For streaming leader Spotify, it was necessary to grow the personalization team. Within a few short years, the company went from having one tiny team to multiple teams located in New York, Boston, and Stockholm.

Spotify notes that platform features are “but the tip of a huge personalization iceberg.” Discover Weekly is one personalization feature that delivers a two-hour compilation of custom-made music to Spotify users, which is “based both on your own listening as well as what others are playlisting and listening to around the songs you love … It’s like having your best friend make you a personalized mix tape every single week.”

Release Radar is a feature that Spotify added in 2016. It provides users with a weekly playlist with songs from new albums. The company admits that for new albums, there isn’t data about streaming or playlists to draw from. Instead, it relies on “audio research” that uses deep learning to curate similarities between music. By comparing this research to a user’s listening history, new tracks that might pique a listener’s interest can be determined.

Pandora’s Big Data Music Project

Spotify’s audio research approach is similar to Pandora’s Music Genome Project, which began in 1999. In this project, teams of trained musicians listen to and categorize music. From that work, Pandora has built a collection of 450 musical attributes that describe a song’s melody, harmony, rhythm, form, instrumentation, lyrics, and voice.

That work informs the platform’s machine learning and algorithms. Pandora combines the data compiled in the project with data about what users listened to, as well as the listener’s thumbs-up or thumbs-down choices when they are presented with a new song. With the Music Genome Project, Pandora hopes to deliver the “most comprehensive music analysis ever undertaken” and provide listeners with “personalized radio that plays what you love and continually evolves with your tastes.”

What Lessons Can Be Learned?

The sheer volume of music and other audio being streamed by listeners reveals that streaming services have tapped into a strong consumer need. This popularity is not something they’re taking for granted. The lessons being applied to big data and music offer insight that other consumer-focused retailers should heed.

  • Give consumers a reason to stay. Spotify grew their music personalization team for a reason: engagement matters. Use data to drive loyalty and inform the features and options you create for customers.
  • Recognize that consumers want to be known by businesses they use. More than 50 percent of surveyed millennials, Gen Xers, and Gen Zers appreciate receiving product and service recommendations based on their purchase history.
  • Understand the importance of being data-driven. Retailers can no longer survive operating on instinct or gut feel. The world is being driven by data, and companies must commit to creating the systems that support their data operations.

Live-streaming platforms continue to explore more ways for big data to drive customer engagement and loyalty. The growth and customer satisfaction being driven by these efforts should be music to the ears of retailers of all kinds.

Personalized data and real-time analytics are reshaping the retail landscape, and it’s time to discover what they can do for your business.

Leave a Reply

Your email address will not be published. Required fields are marked *