Entertainment Intelligence (Ei) is changing the way the music industry collects, views and utilises data.
Not a generic reporting platform, Ei builds tailor-made, ever evolving solutions to help the industry truly understand the market, to save time and to increase efficiencies.
Ei is unique in its ability to ...
Artists who are self-managed, or working with smaller teams or labels, often hear about the data we provide to labels and distributors and ask some variation on the question, “Why do I have to worry about this?” on top of all the other concerns of a working musician. It doesn’t take long to show why data can help them make good decisions, but then they face the next obstacle: where to get access to the data, when so few services provide it in helpful forms, and how to make...
Artists who are self-managed, or working with smaller teams or labels, often hear about the data we provide to labels and distributors and ask some variation on the question, “Why do I have to worry about this?” on top of all the other concerns of a working musician. It doesn’t take long to show why data can help them make good decisions, but then they face the next obstacle: where to get access to the data, when so few services provide it in helpful forms, and how to make sense of it.
These obstacles can be overcome, with just a little care and planning. But before we get to our advice on how to use data as an independent artist, it’s worth breaking down what we mean by “data,” where it comes from, and how artists need to understand it to thrive in this digital age.
Data Is Everything
Don’t worry, we’re not saying data’s all that matters, that it’s omnipotent. What we mean that every online device, transaction, communication, interaction, even emotion, is recorded somewhere, in some form.
Your mobile phone, smart speaker, TV, smart fridge, online food company, train operator, parking lot CCTV, coffee shop, office building, laptop, and email provider all record billions of lines of data about where you are and what you are doing. So before you’ve had time to ask who had a good weekend, there is a data trail of when you got up, what meal you’re planning for tonight, how you feel about the local election candidate, how often you miss your train, how much caffeine you need to feel human, the amount of litter you drop or bin, your favorite parking spot, and the private correspondence you finished off at your desk.
Retailers, influencers, broadcasters, sports teams, churches, even politicians we are discovering, are using whatever data they can get their hands on to attract, guide, encourage, or confuse you into listening to their message, and the same is very much true of the music industry.
No matter how we feel about the way data is being generated, collected, and used, it is happening and is not going to stop any time soon. So, artists, managers, promoters, agents, labels, distributors, and curators all need to learn how to embrace and benefit from the rich vein of music information that is out there.
Data Is Different
Not all data is equal and not all data is important, so knowing the difference is vital to an effective career in music.
The first thing to get straight is that Metadata and Consumption Data are very different things. The first is catalog information relating to a release of music, the label, artist(s), composers, genres, identifiers, street date, copyright details, royalty splits and digital assets (tracks, packshots etc.) that ensure the “work” is properly listed, sold, and paid for.
Consumption data includes all forms of sales, streams, downloads, video views, radio spins and live or public performances of the work. This can range from a few record shop sales of a deluxe vinyl album to 10 million YouTube views of your latest single, from an in-store performance to 100 superfans or 10,000 streams on a Hot Hits playlist.
Just to be clear, when someone says they are working with “millions of lines of music data,” it’s probably just metadata. If it’s billions, then it’s consumption data, from which you can determine the size of your audience, their demographics, location, preferred devices, favourite playlists and daily listening habits.
Data Has Value
Anyone who says they can guarantee to pick the next hit, just from data alone, is selling you snake oil, and anyone who says they don’t need data to inform their gut feel about a band is a fool (one winning pick lets them gloss over all the ones they made that failed). But the two forces together, that combination of guts, determination, patience, endless practice, and meaningful information, can allow you to make excellent decisions and potentially thrive in the music industry.
Let’s look at a simple example of how data can change the way we approach a decision, in this case what playlist to pitch hardest on Spotify. In the following example, we can see two folk playlists with a slightly different average track streams per day (ATSPD) benchmark. This is calculated from the average number of streams each track generates per day while on a playlist.
Fresh Folk (see figure 1) generates a decent number of streams per day. However, tracks are cycled through very quickly, so on average a track can only expect to be on here for 7 to 14 days (its lifespan). If we multiply the ATSPD 1,625 by the maximum track lifespan of 14 days, you can expect to get about 22,750 streams from your time on this playlist.
If we compare this to another popular playlist in this genre, Nordic Folk (see figure 2), we can see the ATSPD of 504 is less than a third of Fresh Folk. This number makes this playlist seem less promising. But here’s where it gets interesting: The retention of tracks is a lot higher, at about 380 days.
So, if we do the same calculation, it’s easy to see the return on this playlist is over 190k streams, which makes it worth fighting to get on to.
Both flavors of data can unlock similar insights that help you focus on what matters most, in a sea of possibilities. Here are just some of the ways you can improve your metadata and do more with your consumption data.
Keep Being Creative
All this talk of metadata, audience demographics, playlist churn rates, and the like may make your head spin. In truth, some music analytics reports feel like you need a PhD in data science to understand them. It shouldn’t have to. You should focus on your music, not mastering statistics.
Here’s where you should press your team, be it an indie distribution platform or manager or small label, to help you get data clarity. At Ei, we serve up the information to artists to make informed decisions and push ourselves to find the right level of granularity. We listen to our clients and work hard to present what we know in an understandable and usable way. You need to make sure your pick a team that does the same, so you can focus on being the amazing creative force you are and spreading your art to those who appreciate it. Data should help, not hinder you, in this, letting you respond to your fans and predict your potential audiences in an informed way.
Charts are nice if you’re part of music’s top 10%, but what about the other 90%?
Public data analytics dashboards look cool, but what do they really tell us?
Music data analysis experts Entertainment Intelligence (Ei) have established a whole new approach to help artists and labels see where they stand with fans--and to guide them to better marketing and creative decisions in a global music market. Using a large body of granular data, it has generated a powerful set of Indie Benchmarks (iB) that offer actionable insights to independent labels, artists, agents and marketeers on par with the internal tools used by the majors.
“Charts are an artefact of the physical retail era, when we only really cared about how many units were moving out of the store,” notes Ei founder Greg Delaney. “Data like total overall sales or follower counts are of extremely limited use; how many people go and unfollow an artist or playlist for instance?”
Ei uses listener-level data direct from DSPs like Apple and Spotify, anonymizes them across its client base which includes labels like Concord, Domino, and the Secretly Group. They then apply machine learning to uncover relevant audience patterns and peer artists, creating truly useful benchmarks for independent music professionals.
As in all data analytics, these benchmarks are built on a clear understanding of what billions of lines of data say. These benchmarks answer questions like how many engaged fans does an artist need to meet a release goal or to catch the ear of a key curator; how is marketing spend actually impacting listening and engagement; and what playlists will help an act get where it needs to go. Benchmarks identify similar artists to compare performance with. They add nuance and real insights to basic numbers like followers, streams, and skips.
Skip rates may seem straightforward, for example, but the real question is who is skipping and when. Ei can tell you. “Because we store every single stream, including the listeners’ identifier and their streaming history, we can tell how many times fans have heard a track already and if they have saved it,” Delaney says. “So the next time they skip that same track, we can mark it as neutral, which is very different from a first-time listener skipping, especially the sub-30 seconds skips, which we also track.”
Ei already generated an average track streams per day (ATSPD) benchmark for all playlists, but it can now report the expected lifespan of a track. With this, users can estimate the true value of a particular playlist, versus others that have a large number of followers but lower listenership, or high volumes but an even higher churn rate, the speed at which tracks are moved off a playlist.
The world suggested by benchmarks, beyond the blunt instruments of charts, social listening, and basic counts, is full of potential. “There is a massive audience out there for the music artists have put their heart and soul into creating,”
“If you’re a pop diva, stadium rocker, or DJ extraordinaire, you may well appear on top international charts, but why would those have any relevance to a psychedelic goth rock ensemble with a small but fiercely loyal fan base in Chile? Using benchmarked data, an artist could know if they are position 901 and what of relevance is happening 10 places above and below them,” Delaney explains. “A smart system would then provide tips and feedback on how to improve their position, find a larger audience, book better tours, and so on, based on what worked for others in their position. We need to work together, pooling data and experience to create something worthy of the artists and fans we serve.”
It’s fun to talk about data, its ubiquity and importance. It’s not so fun to make sense of a billion lines of data from multiple service providers for an entire label roster.
That is unless you’re Greg Delaney, the founder and driving force behind data analytics platform Entertainment Intelligence (Ei). The service has been quietly helping independent labels and distributors work with data like the majors do, analyzing playlist performance, setting meaningful benchmarks, and tracking spikes that pop out of deep catalog. Clients include Domino, Secretly Distribution, Sub Pop, Epitaph, Naxos, Concord Music Group and Zebralution, among others.
“If the indies all get together, you can have the tools that rival, if not surpass, what the majors can do,” Delaney states. “It’s hard for a big company to pivot quickly, but smaller labels can, with our support”
After founding Crowdsurge, a global fan-club ticketing company that worked directly with artists like Paul McCartney and Foo Fighters, Delaney wanted to address a different “data mess” he saw in the music industry. He teamed up with label and publishing veterans to home in on the metrics that matter, broadening data’s role from short-term blip-like guide in promo campaigns or blunt A&R instrument, to a finely honed strategic tool in the long game of artists career development.
“There are many DSP dashboards out there,” explains Delaney, “and just as many services offering to unite them into a dashboard to rule them all. Most are scraping limited public data and then offering vanity metrics that aren’t very helpful at guiding business decisions. We discovered labels need nuance and actionable benchmarks that only listener-level streaming data can provide.”
Ei gets permission from clients to directly access their data from digital service providers (DSPs like Spotify, Apple, Pandora, and Deezer). It then anonymizes and aggregates this across clients to generate high-quality insights about playlist performance, source of streams, and listening behavior. It pings catalog owners when it detects a “heartbeat,” such as an older track picking up steam. It allows for music professionals, managers, and artists to do something retailers have done for years but the music industry hasn’t: detailed cohort analysis. Ei also gives important context to data and reports, giving a per-capita option for territory statistics, for example, revealing when a market is punching above its weight.
Together, this renders a richer portrait of just how listeners around the world are interacting with music, a view that goes way beyond raw follower counts or skip rates. For example, if a label sees an uptick in streams around the announcement of a GRAMMY nomination, they can determine how many are new listeners, how these listeners interacted with tracks (Did they move on quickly? Did they add it to a personal playlist?), and where they came from. If a touring band needs to know which mid-sized city to pick in a region (Buffalo or Rochester?), Ei can help figure out where a more likely audience lives, right down to the neighborhood. If a label is considering which playlist to pitch a new single to, Ei can give insights into how many total streams are likely to result from a placement, based on the average track streams per day multiplied by the average “lifetime” of other tracks on that playlist.
“Most people don’t unfollow a playlist, which means followers are not a useful metric,” Delaney notes. “We wanted to find something useful, so we took data from a wide range of genres and around 350,000 artists and four million tracks with worldwide audiences. We used it to create our own benchmarks, something only really big volume folks like the majors can usually do.”
This data analysis yields fascinating and actionable results. A large indie label noticed something strange happening in several East Asian markets with a particular song. Listeners in Japan and Korea all seemed to be skipping at the exact same spot. Meanwhile in Central Europe, listeners kept streaming away. The label found that right at the skip spot, a harsher guitar part broke into what was an otherwise chill song. This observation inspired a remix that changed that transition, tailored to those markets.
This kind of data-driven approach is particularly suited for the way indie labels work, to the longer-term investment in artists and repertoire they make. To really see what works, they need to cultivate the wild data pouring out of DSPs.
“Wild data is out there, and it can run free and evolve,” reflects Delaney. “In walled gardens, the data will never evolve, and we’ll never get better guidance. We need to let the data out of all the proprietary silos and private databases, let it cross-pollinate and evolve. By working together we can give everyone in this industry the power to make data-informed decisions.”