Exploring What's Coming Next for radio broadcasters with streaming competition.  

While earlier generations may have turned to FM radio and MTV for exposure to new sounds, TikTok and Spotify have become channels for music and discovery. With listeners embracing on-demand, personalized content from platforms like Spotify, how are we listening to hits now? 

Radio could be a hitmaker again. 

After years of listeners flocking to streaming services for their music fix, radio might be about to get a second act.  Several recent deals and acquisitions suggest that at least some of the more prominent broadcasters see an opportunity.  They are snapping up smaller stations and adding others to their portfolios to improve music programming to better compete with streaming platforms. 

Radio is coming back as people look for ways to entertain themselves without screens and technology, says Mark Fratrik, chief economist at BIA/Kelsey. 

Mark Fratrik said it is something you can do while doing other things. It has a shallow barrier to entry. You don't have to sit down and watch it. The experience is also communal, which means it could appeal to advertisers looking for a way to reach audiences beyond social media.  At MusicDatak, we think there will be a resurgence in radio audience and advertising because of its ability to reach mass audiences and local communities. The question isn't whether or not there will be growth in radio; it's where that growth comes from.  We believe that shift will come from radio broadcasters using data from streaming platforms and social media channels to enhance music programming and content. 

Spotify & TikTok's algorithms are changing the music industry.

While Spotify is a music streaming service, you can use its techniques to follow listening habits to track listeners for other benefits.  Listener Segment Profiles allow you to quickly determine which listeners may be interested in a particular product or service and provide information about how those listeners think and act.  The result? Better targeting of listeners and higher conversion rates. 

Spotify's Algorithm Knows Exactly What You Want to Listen To.

There are about 200 factors Spotify considers when deciding what song should play next. While Spotify is known mainly as a streaming music service, they also have business clients who use its database and algorithms to get more information about their listeners' listening habits and preferences to create unique marketing campaigns or target listeners with advertising opportunities based on what they think that listener will like.  The result is that each time you listen to music on Spotify, you contribute to their vast database of information about human behavior, even if you don't realize it. After you play a song, Spotify's algorithm uses an average of about 200 different factors—from how long you've been using Spotify and what songs you play to things like where you are in the world, when you listen, and how many devices are nearby—to determine what song will be your next music hit.  It's more complex than just choosing a song with similar characteristics as ones you've listened to before; Spotify looks at many factors that affect each listener differently.  Spotify can change which factors it takes into account based on who is listening or what they're doing at any given time, making every session unique. And all those algorithms serve as one component of some pretty impressive personalization software.  

Connecting listeners with songs they'll love.

A set of powerful algorithms uses your music and details to shape your entire listening experience. Using everything from anonymized metadata like gender, age, location, genre preferences, artist relationships (including collaborative projects), playlist participation history, favorite artists/genres/danceability/etc., offline mode usage patterns, data from third-party partners (e.g., Facebook), ... Spotify matches listeners with specific recommendations via personalized editorial lists and listener-curated playlists that are tailored specifically for them.  The hope is that by connecting listeners with songs they'll love—rather than giving them random tracks off someone else's playlist—they'll stay on Spotify longer and keep coming back for more. But what happens if a recommendation isn't quite right?

Spotify looks at a vast number of factors that affect each listener differently.

Whether you're listening while doing one of your usual activities or just relaxing, Spotify's algorithm adapts itself to a few things: activity, genre, BPM, and valence. In other words, as is true with everything in life, context matters.  Valence refers to whether a song is described as happy or sad. Danceability refers to how hard it is for someone listening (probably you) to keep still. The higher it is, generally speaking, the harder it'll be for you not to tap your foot along with a song.

How Do Spotify's Unique Matching Algorithms Work?

The company's algorithms pair listening habits with specific traits and predilections. It is possible to conclude a person from their musical taste. Music can determine other attributes of your personality.  Certain songs will naturally put you in a good mood or evoke strong feelings based on your personal history.  Spotify uses these patterns to match each listener's listening habits with certain human traits, then stores that data in a Listener Segment Profile, USP.

Why does Spotify Decide To Look at The Link Between Playlists and Personality?

When Spotify initially decided to examine how listening patterns correlate with personality, one of our main questions was, what's different about people who like a specific type of music?  After analyzing millions of playlists, it became clear that Spotify listeners fell into distinct groups. Each group had several playlists in common; these commonalities gave us insight into each group's collective personality. Spotify's five-listener segments (Curious Urbanites, Party Animals, Hip Hop Hounds, Exercise Junkies, and Sleepyheads) represent more than just music preferences.  By understanding what kind of person listens to a specific type of playlist—namely which traits they possess—we can start predicting listeners' behavior, motives, interests, and needs.

What Did Spotify Research Show Us About the Link Between Playlists And Personality?

We thought it would be interesting to dive deeper into Spotify's USP data to discover what listeners have in common.  After many number crunching, we generated surprising insights about music and human behavior. For example, we found that people with classical music playlists tend to be perfectionists, while those who listen to live sets prefer getting things done ahead of schedule.  While these results are not comprehensive or definitive, they shed light on how different personality types like other kinds of music.

What Are the Links Between Genres And Personality?

There are strong correlations between specific music genres and human personality. For example, people who listen to Dance Music are likely extroverts; fans of Country Music tend to be agreeable; those who like Pop lean toward neuroticism.  The data shows that listening habits can reveal far more about us than we'd think possible. Spotify built a USP based on our listeners' listening habits that can help marketers develop hyper-targeted marketing campaigns.  This has obvious value for B2C (business-to-consumer) companies and proves valuable to B2-B (business-to-business) firms.

What are the Links Between Moods And Personality?

We found that different songs correlate with other traits. "Hallelujah" by Leonard Cohen and "Little Lion Man" by Mumford & Sons show a positive correlation to words like "tremendous" and "beautiful," while pop songs like Katy Perry's "Firework" correlate with more negative traits like "terrible."  The data also showed that our listening habits change once we are primed to feel a certain way (say, upbeat). For example, people will listen to more joyful music afterward if you play a happy song first (like Katy Perry).  We saw something similar happen if you primed people with sad music—they started listening to more melancholy songs later on. Spotify's primary goal has always been to help listeners discover new music and enjoy their favorite songs as much as possible. They accomplish this through AI (artificial intelligence) and an open API (application programming interface).  For example, the Discover Weekly playlist determines precisely what each listener wants to listen to by drawing from their listening history and personal profile data. You'll probably enjoy similar artists in your Discover Weekly playlist if you like hip-hop. 

The Three Parts of The Spotify Algorithm.

Spotify uses several different formulas to track what music you listen to and predict what music you want to listen to. Two of those formulas work together to mimic a traditional radio station—and they make much sense when you think about it.  There's your musical DNA, which is based on what songs you already know and like; there's your play history, which is based on what songs you click play on; and there's your Discover Weekly, which is based on machine learning that sees your habits over time (i.e., your musical DNA) and predicts whether or not that means you'll enjoy new music. 

Discover Weekly

It all starts with Discover Weekly, a personalized playlist of music hits and recommendations from Spotify. The service creates an algorithmically customized playlist for each listener that changes every Monday.  This list is your introduction to new artists based on your previous listening habits. Discover Weekly uses information about your likes, age, location, gender, and past listening history to determine what songs you might like. It's like having a personal radio station programmed by an expert DJ just for you.

Release Radar

Every day, Spotify curates a customized list of new music releases that it thinks you might like. But don't think of these as new music releases in the traditional sense—these are songs and albums that Spotify has identified as being within your listening tastes.  So if you like indie rock, you might see an album from Best Coast or Speedy Ortiz; if you're into pop music, Top 40 artists could appear on your radar. Any time a new album gets released by anyone who has similar musical tastes as you do, it could end up on your Release Radar playlist.

Daily Mixes

Spotify creates personalized playlists based on your music tastes, plus it makes a daily playlist called Daily Mix that recommends new music based on what you've listened to before.  These mixes are an excellent way for music fans to discover contemporary artists and songs they might not have heard. No matter what kind of music you prefer, there's a good chance Spotify will be able to recommend something new and exciting.  The company also has several radio stations that play similar artists or focus on a specific genre or decade.

Music And AI In TikTok

With the growth of TikTok, Algorithms and AI are playing an increasingly important role in the content posted on the app.  TikTok has created two different categories of AI to curate content, one that uses computer vision to detect videos and one that uses natural language processing to identify popular keywords in videos.  These algorithms allow TikTok to serve up more and more relevant content to its listeners, but the company isn't planning on stopping there.

Deep Learning

The latest crop of music-intelligence algorithms leverage deep learning to process video data in real-time.  That means there are more potential hits than ever before - especially on platforms that give artists direct access to new ideas through AI. Content creators can now search, find, and incorporate music into their videos in inconceivable ways a decade ago.  What's next? How will AI change how we listen to & create music? And what does it mean for audiences who love watching these videos? 

A New Musical Frontier

At its core, understanding human music is vital to any AI-powered music app. Data is so essential for TikTok—with over 1.5 billion uploads from people worldwide, TikTok uses data to find patterns and teach our systems how people make their music.  Then, we can use those patterns to make future posts better! For example, if a particular hashtag is popular in one country but not another, TikTok might suggest it to listeners.  Or, if a specific type of meme is trending on TikTok in one region, TikTok might recommend it more frequently. It's easy to see how these trending topics could be a starting point for generating new memes or even new genres of music!

How Music Is Used On TikTok.

Like other social networks, music connects listeners & content. However, unlike its predecessors, a piece on TikTok is designed to transcend particular genres & age groups; tunes run up against other tunes in a continuous stream of *content*.  This way of thinking about music changes how we consume it: instead of enjoying hits from specific eras (or simply listening to our favorite songs), audiences are exposed to variety via an endless feed of songs (popular and obscure).  It also allows for new forms of collaboration between artists: as dance challenges or lip sync videos become popular, others quickly adapt them—and then riffed upon them—creating a musical feedback loop that keeps listeners engaged. 

Musicdatak is the path to discovering the hits your audience is listening to and all the data that comes with it.