By Mitchell Shymansky
Today, Universal Music is launching a tool created specifically to inform and empower our artists and their teams. Called Universal Music Artists (UMA), it will:
· provide data to those that need it on-the-go, in a beautifully designed and expertly crafted application available for both iOS and Android;
· for the first time across all major multi-territory music services, show artists what their audience looks like, where they are located, how they are listening to music and how frequently; and
· provide artists with data allowing them to gauge the success of their releases and provide actionable insights and suggestions.
At its heart, UMA is all about providing artists with more and better data to deepen their understanding of their fans and their listening habits. Unlike any other artist-focused tool in the market today, UMA uniquely provides a view across major consumption and social media platforms.
To begin, our application uniquely features data from the use of our music on Spotify, Apple Music, Amazon and YouTube, and combines that with social data from the conversations taking place around our artists and music on Facebook, Instagram and Twitter. Over the course of 2020, we’ll add data from Deezer and additional platforms. Ultimately, our desire is to include every platform in every territory around the world to ensure our artists and their teams continue to have the best data and insights in the entire industry.
Of course, data-driven insights aren’t very useful if an advanced degree is required to comprehend and act on those findings. In the words of UMG Executive Vice President of Creative (and former Spotify Global Head of Artist and Label Marketing) Dave Rocco, “you need to make data sexy!” So, we teamed up with YML, a Silicon Valley-based design and innovation agency, to create a streamlined user-interface for UMA and a suite of tools capable of making good on Dave’s directive.
Labels: “laggards” or leaders?
Innovation rarely happens overnight, and, in the case of record labels, there has been a perception for quite some time that they are technology laggards — slow to change, and slow to embrace new methods of thinking. Frankly, while it’s easy to say this might have been true in the past, it is simply not the case today. In fact, I’d argue that record labels have become leaders in the creative use of technology, particularly as it relates to moving the needle on culture and the arts.
It’s said that growth only comes after a little hardship, or a little suffering. Well, given what this industry endured through the early 2000’s, there was a wellspring of challenges to drive growth — shaping us, and fueling an entirely new way of thinking. Change has become the only constant in this industry, and we have had to become experts at adapting and evolving our business just to survive. But in the end, we have not only survived the Internet revolution; we are thriving. Along the way, we provided a successful roadmap forward for other media and creative industries, like newspapers, television, film and broadcast radio.
Today, we constantly use data to innovate and change the way we do business. UMG’s team of data scientists are busy building things like “hit prediction” models, content valuation models, playlists survival rate models, marketing attribution models, as well as developing additional next-generation tools that incorporate machine-learning and A.I. We’re leading the way, and the UMA application will be our outlet to get all of the resulting market intelligence into the hands of those who can act on it. What we launch today is the start of a long-term process. It will only get better and better, until, we have infused it with data from every facet of our business.
Even with this data-driven approach, at the most fundamental level, we haven’t changed our relationship with music, artists and the creative process. We’re still in the business of making music. Whether it’s pop, hip hop, country, jazz, or metal, we help our artists realize their talent and potential — creatively and commercially — and deliver their music to the broadest possible audience around the world.
We will always be emotionally moved by music, and our artists will always be central to what we do. On top of that, we also focus on music fans. Record labels have spent decades anticipating how people would react to their music. Now our data gives us an unprecedented look into what music fans like, and the tools we’ve built allow artists to better connect with their audiences than ever before.
For years now, experts across all industries have been talking about the value and importance of data. How data is collected and used will transform industries beyond all recognition, they’ve said. Today, we have incredible examples of companies using data to optimize their processes in ways that were unthinkable just a few short years ago. Companies’ wherewithal to “crunch the numbers,” so to speak, has very real effects; revenues and profit margins have expanded or contracted based on one’s ability to adapt to this change.
To be successful in this new world, companies must truly and deeply understand their audiences — to anticipate their needs and desires and build engagement by building and acquiring the right products and content.
By now, you’re probably familiar with the term “the attention economy.” In the past, revenue was largely derived from physical media: newspapers, DVDs, and CDs. Today, companies are fighting for “mindshare.” Attention — acquired, packaged and then efficiently monetized — is the predominant business model in the modern, connected world. Entertainment companies have always been in the business of capturing attention, but now we’re using data to more effectively target the right consumer at the right time.
The recorded music industry’s revenue peaked way back in 1999, when Britney Spears, NSYNC and The Backstreet Boys led the charts, and sales growth appeared to be limitless (or so it seemed). At that time, the position of a data analyst within a record label was unheard of. This was mainly a gut-feel game driven by charismatic A&R folks and marketers, supported by a healthy music eco-system consisting of record stores, music magazines and MTV.
But then file-sharing and MP3s led to declining profits, thousands of lost jobs, “the end is nigh” proclamations, and so on. The music industry was turned on its head and destined toward oblivion. The public believed labels were filled with backwards, technologically challenged Luddites, and the game was up.
When I joined Universal Music in 2006, everyone asked me why I would want to work in a dying industry. Not only were CD sales declining, the download market hadn’t been able to make up for these lost sales. Regardless, I was excited by the problems to be solved. I saw an industry that was being forced to “up its game” — and fast — and that was exciting. Labels were changing and evolving, hiring digitally savvy executives for top positions. This was exactly where I wanted to be, because I could anticipate the impending influx of data and the opportunity it represented.
As it turned out, streaming wouldn’t result in a rebound for industry until recently. However, well before then, a fundamental shift began across our entire industry. For more than a century, a record label’s job was to convince people to engage in a series of one-time transactions. But today, with fewer and fewer consumers paying to own physical CDs or download files, we are predominantly focused on persuading people to give our artists and music their ATTENTION, over and over again, each and every day.
More than 255 million people at the end of 2018 paid to access everything — tens of millions of tracks and growing daily — on any one of dozens of premium streaming services, according to the IFPI. With very few exceptions, countries around the world are experiencing incredible growth in paid streaming. For example, last year premium service revenue grew 33% in the United States, 30% in Japan, 23% in France, and 35% in Germany (for both free and paid services).
I offer this brief history of digital music’s early days only to introduce a new business model teeming with data. The ownership (transactional) model could offer only so much data. A single purchase results in a single data point. Retailers used to give us relatively rudimentary sales data that gave no insight into the customer or their usage and motivations. In contrast, the access model provides labels a large and dynamic volume of the most important assets that will drive our future: streaming data. We receive deep insights from key digital platforms every single time a fan presses play. This is a new world for the music industry and UMG is at the forefront of adapting to this new way of thinking.
From intuition to intuition + data
Creatively, the music industry has largely been driven by instinct, intuition, or quite simply, “gut feel.”
But what do we mean when we talk about these things? We’re talking about the subconscious mind’s ability to process signals and cues — data, in other words — in our everyday surroundings. Think of A&R executives standing at the back of packed gigs and getting the sense that something “special” is happening on stage. They’re running data across subconscious mental algorithms, refined over time and from countless prior experiences to form those judgments. As humans, all of our decisions are driven by data, whether we know it or not.
Today, thanks to the large volume of music from UMG’s labels played on Spotify, Apple Music, etc., we are able to supplement our executives’ intuition with real, hard data. We have access to signals that help us understand what our audience looks like (age, gender, etc.), where they are listening (country or region), how they discover and listen to music (playlists, search, etc.) and how frequently they listen.
I like to think of the data and analysis we receive as a sort of pro-biotic for our executives’ well-honed gut instinct, if you will. We’re now able to utilize this data to identify high velocity content, make decisions on how to acquire and promote the best content, on which channels, and for how long. We can use the same data to make predictions on the size of the audience likely to interact with this content, along with how best to communicate or market to that audience.
Simply put, we’re able to push the right content, with the right message, to the right fan, on the right platform, at the right time. We’ve gone from marketing to broad swathes of masses, to operating data driven campaigns with surgical precision. By maximizing our impact and optimizing our spending strategies, we can create a more effective and efficient marketing machine. And, most importantly, we are pushing UMG artists’ careers to new heights.
Making data useable
Having large volumes of data is both a gift and a challenge: it’s wonderful to have but difficult to make functional. We first had to make our data useable.
We needed to turn our fast-growing quantity of data into actionable insights — information you can actually use — rather than provide artists with statistics that have no practical purpose. So, three years ago, we began investing heavily in improving our analytics infrastructure.
We have accrued 1.05 petabytes of data from our music being collectively streamed across only three platforms: Spotify; Apple Music; and YouTube. For some perspective, 1 petabyte is the equivalent storage capacity of 20 million four-drawer filing cabinets filled with text. It’s safe to say our prior infrastructure was just not up to the task of handing this much data.
To help us store and process the sheer volume of information, we rebuilt our entire analytics environment within the Google Cloud Platform. In plain English, this means we have a vast amount of horsepower at our disposal which allows us to make sense of the constant firehose of data that we ingest every single day. One of the many advantages of being the leading recorded music company in the world, with an incredibly supportive parent company in Vivendi, is that we have also been provided with the necessary resources to build and maintain a world-class, internet-scale analytics infrastructure.
We also brought an entirely new skill set into the music industry. We have hired data architects, data engineers, user-experience designers, statisticians, computer scientists and more. Not the usual cast of characters traditionally found within the industry of old. And although it doesn’t sound like it, this is an extremely creative team. I mean, it takes creative genius to turn one quadrillion bytes of data into something that helps people actually make decisions, and fast.
All about artists
Providing our artists and their managers with data and actionable insights has remained a top priority at Universal Music Group for more than a decade, a mission that has been directed and supported from the highest levels of our company’s leadership.
We understand it’s essential to provide artists with better data not only to forge stronger partnerships, but to empower more informed decisions through a deeper understanding of fans and listening habits and to work closely together in the shared goal of advancing careers.
While our data journey started in 2007, when we began collecting monthly iTunes sales data on a computer under my desk, as I’ve explained, we didn’t have access to robust data until streaming began started to take off. Our first attempt to provide artists with this incredible information was a tool called Artist Portal, which launched internally in 2010 and became available to our artists and their management teams a few years later.
However, as streaming began to take off, we understood that we would need a completely rebuild Artist Portal into a tool that could scale with the escalating amount of incoming data and present information in a concise, digestible manner and that could simply surface key, actionable insights.
As we built the new cloud-based architecture that underpins Universal Music Artists, we also began canvassing artists and managers who used Artist Portal to determine which features are most important to them. Artists and their teams told us they want a topline number that tells them whether they’re up or down in a particular week or month. They want to know how their songs and videos are performing both globally and in specific territories. And they want to better understand their fans and what they’re reacting to on social media.
Armed with those insights, and understanding that data must to be digestible, we began working with YML to ensure the user interface clearly communicated those points in a way that is easily digestible, particularly to those unaccustomed to looking at data day in and day out.
The result is Universal Music Artists. However, as much work that has gone into bringing us to this point, we have only really scratched the surface of building this tool. You can expect many more features, data sources and developments to come along quickly.’
None of this would have been possible without the participation and input of artists and managers — and we hope as you use this app that you will also continue to provide us with feedback and help us make this tool even better.
Because at the end of the day, none of this would be possible without you.
Mitchell Shymansky is Vice President of Data Analytics at Universal Music Group, where he leads a team of data scientists, programmers, and engineers in the pursuit of leveraging the company’s vast data resources to create exceptional new opportunities for artists and labels.