Back To All Articles

Behind the Music: Lessons from Spotify & Pandora in the Age of Machine Learning

With the speed at which new technologies enter and transform our daily lives in this day and age, it can be easy for brands to lose sight of the key principles to establishing and maintaining a competitive advantage in the digital landscape.

Katie Cullina  |  November 22, 2017
Behind the Music: Lessons from Spotify & Pandora in the Age of Machine Learning

Sometimes, you get hit with a pang of nostalgia when you least expect it. Like when you inadvertently open Pandora instead of Spotify before a run. This app-y accident left me unconsciously clicking out of Pandora, and later consciously thinking about how Pandora had gone from being a staple in my daily tech repertoire to an app that I had mistakenly opened while looking for something else. When did I so casually desert Pandora and start breaking bread with the enemy, Spotify? Having learned about Pandora’s humble beginnings and inspiring founding mission in business school, I felt I owed it to Tim Westergren to find out what happened, so I set out on the internet looking for answers. What I found led to more global takeaways that I think are relevant for any marketer to reflect upon, so buckle up and get ready to learn through music, folks.

So, what did I find?

First, I found that I’m not alone in leaving Pandora for Spotify. In fact, I’m among the vast majority of digital music subscribers in doing so.

Pandora has gained 5 million users in the past 3 years, compared to Spotify’s 100 million in the same timeframe.

Exhibit 1

Second, I found that there are several important factors that have caused this shift, and that the key operating principles that have helped Spotify win in recent years can be applied to brand and agency digital strategies as well. For ease of reading, I’ve summarized what I consider to be Spotify’s universally-applicable approaches to establishing a competitive advantage below.

Don’t Get Comfortable

  • While Pandora was first-to-market in a new category, its uncluttered existence was short-lived, and the company failed to evolve its music library, algorithms, and capabilities while Spotify was ramping up in the background, ready to take advantage of Pandora’s first-mover complacency. Sure, this is a simple, maybe even obvious idea, but it’s helpful to remember that approaching digital and social brand “refreshes” or “optimizations” in a fixed-mindset way can open the door to missed opportunities and a constant, unrewarding game of playing catch-up.

Focus on Customer Utility

  • Spotify knew what consumers wanted: the playlisting and music recommendation capabilities people loved about Pandora, and over 30 million songs available for instant access at a lower price. While Pandora initially shrugged this off and ignored the clear consumer demand for the latter, Spotify built its low-cost subscription models and attracted an enormous number of defectors at a breakneck pace. In short, Spotify kept its focus on providing maximum utility for consumers, and this focus helped the company earn a critical mass of start-up users that compounded quickly. Likewise, brands and agencies can benefit from maintaining a strategic focus on utility. By maintaining an obsessive focus on giving customers what they want or need, or what they don’t know they want or need yet, brands will be more effective in getting them to routinely come back for more.

Push the Boundaries of UX

  • While Pandora doesn’t allow for users to build playlists and replay or search for songs, Spotify allows users to move freely within the platform, access the service on any device, integrate activity with social channels, and explore whatever charts and playlists they’d like to. Spotify is designed for people to spend time with the app while easily finding what they want to find wherever they are, and the market prefers this architecture based on the demonstrative active user metrics (Exhibit 1). This sort of omnichannel, seamless, and experience-centric approach is important for brands and agencies to keep top-of-mind as they think about inserting a product or service into their target audience’s usage behaviors.

Double Down on Data Science

  • Believe it or not, Pandora’s early recommendation algorithm was powered exclusively by living, breathing musicians. Initially unpaid, they sat and manually coded as many songs as they could physically process according to unusual qualifiers like instrumentation, rhythmic intensity, and tempo. These unique filters in turn enabled the music discovery mission of the company; as users indicated what they liked or disliked, the algorithm better understood the specific musical elements they were drawn to and served up artists or genres they wouldn’t have previously considered. While the quality of execution was impeccable, Pandora’s approach lacked scale, and as the company focused on slowly building its human-powered music recommendation engine, Spotify perfected a machine learning approach to this task, and lapped Pandora in the process. Today, Spotify boasts the widely-praised and machine learning-powered “Discover Weekly” and “Daily Mix” customized playlist offerings, and demonstrates the fact that a laser focus on data science innovation and evolution is increasingly critical to establishing and maintaining market dominance.

Underlying all of the aforementioned points is the importance of timing. Spotify entered the market in the right way and at the right time, and it’s unlikely that Pandora will be able to chip away at Spotify’s market stronghold in the immediate future. While we can’t necessarily replicate these market dynamics for brand campaigns, we can certainly learn from Spotify’s market instincts, constant evolution and laser focus on providing seamless and valuable user experiences.

Back To All Articles

4 Headquarters Plaza, Morristown, NJ 07960

©2022 Revolution Digital  | Privacy Policy