How the Smartphones changed music and the music industry [Kyle Bylin]

Kyle Bylin explores how the smartphone and the apps that followed have reshaped and continue to reshape music and the music industry.

by Kyle Bylin

SEVERAL WEEKS AGO, I found a book in Barnes & Noble called Computing Taste: Algorithms and the Makers of Music Recommendation by Nick Seaver that inspired me to reflect on my career and interests, especially those that might be interesting to explore in a Ph.D. thesis. I’m currently a graduate student at the School of Information, University of Michigan, Ann Arbor, pursuing a master’s degree in information science. My studies revolve around several subject areas, including public libraries, user research, community engagement, and startup clusters. Ann Arbor is one of the most educated cities in the United States, and the idea of pursuing a Ph.D. thesis during graduate school is quite prevalent. As a leading research institution, the University of Michigan exposes its students to hundreds of academic papers, making it easy to consider writing one yourself. 

The other day, I wrote a lengthy email to Nick, whom I’ve known since he started working on his Ph.D. in Cultural Anthropology. In the email, I shared insights from my 15 years of experience in the technology industry and my deep passion for discovering artists, listening to music, and using technology to enhance my listening experience. I also discussed possible subjects that I could explore in a Ph.D. thesis or book. I introduced Nick to some of the people who are featured in his book, and my second essay collection, Promised Land, is cited as one of the sources in his book’s bibliography. This made me realize that I am already cited in real books as a source. Therefore, I’m considering writing my own history book on the streaming age. It’s something I’ve wanted to do for years, but I never had the plotline nailed down to the point where I could justify writing an entire book.

I recalled too many Amazon book reviews criticizing authors for writing books that could’ve been magazine articles. Therefore, I was sensitive to the idea that not everyone needed to bind their worldly thoughts into a dead tree; there were other ways to prove that you mattered and existed in the broader and darker universe. 

Pandora’s Music Genome Project: The Evolution of Personalized Music Recommendations

Earlier today, I remembered my desire to write a book about the impact of smartphones on how people hear, discover, and experience music. However, I realized that the title I had in mind is incomplete because we’re now transitioning from the age of the smartphone to the era of the intelligence-enabled device. 

The unsophisticated assistants we once used, such as Apple’s Siri and Google’s Assistant, will be replaced by all-knowing and conversational assistants that can perform a variety of tasks. These tasks include analyzing and summarizing the world’s knowledge and information into any conceivable format, booking a plane ticket, and planning a last-minute vacation based on your user preferences. 

Furthermore, these assistants will gather an immense amount of data to decide which music to play and when based on the popularity of the artist or song on social media at that moment. It’s only a matter of time before artificially intelligent DJs become commonplace and we become accustomed to summoning AI artists, DJs, or fans into the stream whenever we want to discuss about music.

As I contemplated the evolution of music further, I realized that a significant change was on the horizon. This revolution can transform how people discover new artists, listen to music, and watch related videos. The advent of smartphones brought about a couple of transformative advancements that changed the way people experienced and found music. However, the most significant breakthrough came with the release of Pandora’s app, which offered customized radio stations. 

Pandora’s mobile app enabled users to enjoy their customized radio stations on the go, offering a unique music experience tailored to their preferences. By giving a thumbs up or down rating to each song, users could indicate their level of liking for the recommendation. This process may appear simple, but it actually required years of hard work by musicologists and pop culture experts who analyzed every song, identifying hundreds of descriptive variables. These variables were then incorporated into the metadata, enabling personalized recommendations.

For instance, a song such as “Screaming Infidelities” by Dashboard Confessional could be followed by “Cute Without the ‘E’ (Cut from the Team)” by Taking Back Sunday, which could then be followed by “Ocean Avenue” by Yellowcard. This personalized recommendation does not result from a magical algorithm that understands your preferences better than you do. Instead, it’s the outcome of decades of tireless efforts, determination, and commitment by teams working to teach computers how to comprehend music and provide recommendations that keep users engaged with the platform, listening to music, and watching ads.

Shazam and SoundHound: How Mobile Identification Apps Reshaped Discovery

During the smartphone era, Shazam and SoundHound introduced music identification technology that could listen to the music, identify the song’s name and album title, and provide the listener with real-time information. Earlier, this technology existed on other mobile, handheld devices as a 1-800 number, which could be called to identify a song. However, the mobile app was the first product that packaged this technology to solve a genuine consumer problem. Listeners would hear songs on the radio but could not quickly determine the song’s name. 

Shazam and SoundHound solved the problem of identifying millions of unknown songs worldwide by developing mobile apps. While both apps pursued different strategies to keep people on their platforms, retaining customers after recognizing the song proved challenging because they already had the needed information.

Both Shazam and SoundHound faced a challenge to optimize their business models and remain competitive. To stay afloat, Shazam explored innovative advertising strategies, such as mobile app identification with next-generation QR codes. Meanwhile, SoundHound spent over a decade developing advanced voice technology to rival Apple, Amazon, and Google in the Voice Assistant market. 

Eventually, Apple acquired Shazam, which is now a key feature in Siri.

Meanwhile, SoundHound became a publicly traded company, attempting to be revalued for its voice technology efforts rather than the music technology innovations that got it to that point in its product development process. Using your phone to identify a song was revolutionary, but now we take it for granted. 

Today, we listen to most music on demand and always know what song is playing through our headphones. This change is the second most significant innovation in the history of the smartphone, and it has transformed how we discover new music.

Music Concierge: The Rise of Context Culture in Streaming Apps

The algorithm or recommendation engine is one of the most significant innovations that has changed how people experience music on connected devices. It’s the foundation of most people’s experiences with music streaming services today. The considerable difference between Pandora’s music genome project and modern music recommendation engines is that Spotify acquired a music intelligence firm called the Echo Nest. This firm scoured the internet for data about songs and leveraged that information to build a massive network of associations that could be used to make custom recommendations and playlists. 

The Echo Nest could measure the popularity and velocity of a given song in terms of how it was received online. Algorithms like these underpin popular playlists like Spotify’s “Discover Weekly” and “Release Radar” and most personalized playlists and radio stations on the streaming platform. Spotify’s new artificially intelligent-powered DJ, which we will discuss in another essay, also relies on the algorithm.

The music industry has been revolutionized by improved recommendation engines, which allow us to access a virtually infinite stream of personalized songs from our connected devices. Although the initial innovation of music streaming platforms provided unlimited access to millions of songs and the ability to access them anywhere, it proved too overwhelming for the mainstream market. To make the music experience more relevant and personal, music companies turned to recommendation engines that enable users to create customized music playlists. 

In today’s world, music listeners can stream their favorite songs from anywhere. Thanks to the availability of connected Bluetooth devices and smart speakers, listening to music has become even more convenient. The algorithms used by these streaming engines have made the music experience smoother and more efficient by reducing choice overload and enabling people to enjoy their favorite tunes quickly and easily. These algorithms have revolutionized the music industry and have proven that streaming can be successful even in the mainstream market.

The history of music streaming via mobile music apps cannot be discussed without mentioning Songza’s Music Concierge feature. Its unique approach to understanding how people think about music during a listening session sets Music Concierge apart. The feature considered various factors, such as the user’s specific mood and the time of day, to select playlists that would appeal to them. 

The co-founder’s father, who worked in marketing research, advised the Songza team to conduct extensive user research. They developed the Music Concierge interface as it aligned with the mental model of how people think about music but couldn’t express it easily. This interface revolutionized the music experience design and became the most imitated feature of any app, besides the original idea for personalized radio and playlists. Other apps like Spotify, iHeartRadio, and Slacker Radio copied this design and incorporated it into their apps because it resonated with how people listen to and discover music at a fundamental level.

The rise of the context culture was a significant development in the history of music streaming, as it led to the integration of real-time data from different sensors into connected devices, making music streaming more personalized. 

From Liner Notes to Live Lyrics: A Cultural History of Textual Engagement with Music

In recent years, live lyrics have become a standard feature across most music streaming platforms. However, SoundHound stands out as one of the pioneers in offering this feature, enabling music enthusiasts to sing along with their favorite songs without worrying about not knowing the lyrics. Integrating live lyrics into music streaming services marks a significant breakthrough in mobile music apps. 

In the past, lyrics used to be an essential part of buying an album, but it wasn’t always guaranteed that the book would include the printed lyrics. This could be frustrating for music fans who want to sing along. However, with the rise of the internet, finding lyrics online has become easier with websites like AZ Lyrics and Song Meanings. Now, Spotify offers live lyrics as a standard feature on their app, enabling anyone to look up lyrics and sing along with the music. This has led to impromptu carpool karaoke sessions, but not everyone wants to hear them. 

The way we listen to music has evolved over the past few decades, especially with the advent of smartphones and music streaming services. These technological advancements have also led to innovative advertising technologies such as music data analytics, which can analyze the listening habits of music fans to create personalized profiles. This data can be used for marketing and selling products to streaming listeners. Most technology companies that offer music streaming services, such as Spotify, use subscriptions or advertising to generate revenue.

For instance, Spotify has convinced people that music is worth $9.99 a month and that they can tolerate ads if they want to listen for free. However, this essay does not cover the use of music and demographic data for targeted advertising. 

Nevertheless, it is noteworthy how personalized music has led to personalized advertising, which is now ubiquitous in modern life. Bands can use hyper-specific tactics to target their music to specific demographics and musical preferences. 

While the development of the infinite stream and the rise of algorithms in creating playlists that never end have transformed how we listen to music, mobile and live lyrics have also played significant roles in enhancing our music experience.

In the past, record labels faced significant challenges when targeting their marketing efforts towards new artists. Their options were limited, and they relied on traditional methods such as broadcast radio and TV campaigns to reach their desired audience. For instance, if the label wanted to promote a new album for Linkin Park to American men aged between 15-35 who watched the World Wrestling Federation on Mondays from 7 to 10 p.m., they had to showcase a music video during the program that followed wrestling. However, there was no guarantee that the music video would effectively reach the target audience, and targeted ads were not a reality back then. Labels had to rely on mass marketing to promote music, which posed huge cost and audience engagement challenges.

The Rise, Fall, and Resurgence of Short-Form Musical Content: Vine, TikTok, and YouTube

Smartphones have also transformed the music industry by providing new ways for people to consume music videos. YouTube has been leading the way in this change, enabling creators to earn money by producing music videos and allowing unknown artists to become overnight sensations if their videos go viral. 

In the past eighteen years, YouTube has become integral to the video culture ecosystem and creator economy. Many artists now make a living by producing cover songs or performances, while others use the platform to promote their original music videos. YouTube has become a distribution platform that helps people discover new artists. To keep up with the success of TikTok, YouTube introduced a short-form video service that allows videos to be found in new ways on mobile apps. TikTok has revolutionized the U.S. music industry in the last six years, enabling artists to connect with fans, live stream shows, and directly communicate with their audience. It has completely altered how people consume music and opened doors for unknown artists to achieve overnight fame. 

The popularity of short-form videos can be traced back to Vine, which allowed artists to gain viral fame and establish themselves as powerhouses on YouTube.

Vine was a popular app until Twitter acquired it, which caused it to lose its appeal. This created an opportunity for Musical.ly to fill the gap in the market. It allowed young people to create lip-sync videos and helped artists gain popularity in a younger demographic dominated by teens. Later, Musical.ly was purchased by TikTok, which led to an increase in the platform’s musical content and helped its parent company, Bytedance, become the social media giant that it is today.

Users can share songs, playlists, and stories on social networks. Various groups have tried to create music versions of Instagram by combining previously famous music technology startups—for instance, Turntable.FM allows group listening, while Soundtracking combines music and photos to create a new storytelling app. 

In early 2010, the internal innovation lab at Live Nation Entertainment considered launching an Instagram-style app for concerts. The app would have allowed concertgoers to share their concert experiences with others who are interested and create a music-focused version of Instagram that could compete with the social media giant. A number of attempts were made to develop social, audio, and video-sharing startups that would allow people to capture concerts in real-time and share multiple angles of the concerts or their perspectives with their fans, friends, followers, and family. Unfortunately, none of these startups reached a maturity point where they achieved a sustainable business model. Only a few are still operational today, trying to make something out of the live experience.

In the wake of the COVID-19 pandemic, many startups emerged, aiming to enable the live streaming of concerts via web and mobile devices. However, many of these startups have since shut down or gone quiet after being acquired. This is mainly because people can now attend concerts in person, resulting in declining demand for streaming music through their smartphones. While there will always be a sizeable market for live streaming of concerts, whether people opt for virtual experiences over the real thing remains to be seen. Moreover, the advent of augmented and virtual reality technologies, such as the Meta Quest headset or Apple’s Vision Pro headset, can transform how people experience live music. 

Such innovations could enable individuals to attend concerts from the comfort of their homes and feel as if the artist is performing right before them as if they were in the same room. With technological advancements, individuals could experience a Kiss concert with a guitar solo happening as if it were happening before them or feel like Taylor Swift was shaking it off right before their eyes during a live show.

From Algorithmic DJs to AI Maestros: The Future of Music Curation in the Age of AI

Over the past few decades, significant technological advancements have transformed how people discover, listen to, and share music. The introduction of smartphones has made it possible to stream personalized, on-demand music through apps like Spotify and Pandora. These platforms have revolutionized the music industry with their Music Genome project and comprehensive discovery algorithms, enabling users to explore new music that aligns with their tastes.

Music recognition technology, such as that used by Shazam and SoundHound, has made it possible to identify a song even when its title and artist are unknown. 

Songza, for example, offers contextual recommendations based on the time of day and mood, enhancing the alignment of music with daily life. Platforms that integrate live lyrics have also been a significant innovation, allowing fans to sing along with their favorite songs. Short-form video apps like Vine, TikTok, and YouTube’s Shorts have transformed how people discover and consume music content. As technology evolves, voice assistants, AI-powered music guides, and DJs will play an increasingly important role in music curation and discovery. 

The music industry is utilizing advanced algorithms and data analytics to provide more personalized music recommendations to individuals. The industry is creating more targeted and effective marketing strategies by leveraging fan data and analytics. To stay relevant and keep up with the changing landscape, the industry must continue to adapt and innovate, creating new and exciting ways for people to experience and engage with music. The trend is to use technology to create more personalized, context-aware, and interactive social music experiences. 

Although the specific music apps and features will undoubtedly continue to evolve, the desire to use technology to enhance how music is discovered and enjoyed remains steadfast. Music has been an integral part of human culture since prehistoric times, and it’s no surprise that people continually seek new ways to harness innovation to serve this fundamental human need. The progress of technology and new possibilities have given machines many chances to improve the human experience of music. This will lead to a revolution in music during the age of artificial intelligence, which will completely transform the music industry.

Artificial Intelligence and the Musician’s Dilemma: Democratization or Displacement?

Over the next five years, the music industry will experience a revolution in music creation, discovery, distribution, and monetization, all thanks to artificial intelligence (AI) and machine learning. With the help of companies like Anthropic and Google, AI-generated music will allow personalized, on-demand music creation tailored to individual tastes and moods. AI systems will create original songs, instrumentals, and remixes that sound like human-created music. 

Advanced AI algorithms will personalize music discovery and recommendations based on mood, location, activity, time of day, weather, and current events. 

Intelligent assistants have the ability to converse with users about their musical tastes and interests and then curate custom playlists and radio stations based on these preferences. As a result, music will become more personalized and context-aware. Thanks to AI technology, music production will be democratized and available to everyone, even those without formal training. With the help of user-friendly apps, amateur creators can produce professional-quality tracks and beats by simply selecting the desired genre, mood, tempo, and instruments. This will expand music-making beyond traditional players, making it accessible to all.

Advanced natural language processing is set to power a new generation of intelligent assistants, such as Anthropic’s Claude, that can engage in honest conversations about music. With the help of AI, users can ask questions about music theory, interpretations of lyrics, concert recommendations, and much more. 

In the next 3-5 years, record labels and management companies will witness a massive upsurge in voice-controlled music enabled by conversational AI. We can expect a rise in targeted fan engagement and marketing thanks to AI-based music data analytics. Record labels will leverage machine learning to identify super fans based on various attributes, such as demographics and online behavior. They can directly engage with them by doing so since these fans’ fandom and income have become increasingly critical for diversifying the modern artist’s revenue streams.

Music streaming services will use AI to predict users who are likely to cancel their subscriptions in order to reduce the rate of subscriber churn. Music distribution platforms will use reinforcement learning to optimize the placement, pricing, and bundling of songs, maximizing royalty revenues. AI-guided testing will determine the optimal release dates and promotional strategies. Additionally, AI-adversarial networks could be used to create new album/single cover art and music videos.

The potential of AI in the music industry is immense. However, specific challenges must be resolved before it can be fully utilized. These include issues related to copyright, royalties, transparency, and ethics. AI technology may replace some human jobs and creativity in the industry. Nevertheless, AI has the potential to bring about a new era of innovation, accessibility, and personalization in the world of music. Even though it is difficult to predict specific applications of AI in music in the next decade of change, the opportunities and possibilities are infinite.