I recently published a piece about how A&Rs use data to scout and evaluate artists on Cherie Hu’s Water & Music publication, and I wanted to share a bit of the background research I conducted to write this piece. Here’s an exclusive interview I did with Chaz Jenkins, Chief Commercial Officer at Chartmetric, where we discussed the role of data in A&R and how it evolved over the years with the advance of analytics tools. Before joining Chartmetric, Chaz had previously founded Grammy Award-winning labels, an artist management company and was VP International Marketing at Universal Music Group.
Julie Knibbe: Given your experience at Chartmetric and previously as an A&R person, how is data transforming A&R departments?
Chaz Jenkins: First thing I’ll say is something really controversial. I’m actually allowed to do that because I used to be an A&R person. A&R has always been about data. Data has always been a really key component of any A&R person’s job, but we never really think of the data we used to use as “data”. It was just insights, information and acquired knowledge from the marketplace. A&R people have always been able to absorb a lot of information. The more information that they could absorb, the better decision making they could make. If you don’t understand the marketplace, then, yeah, you can still hear some music and think, wow, that’s amazing. But how can you make a good business decision about whether you should sign up artists, or whether you should put that artist together with that producer for example? You can’t do any of this unless you actually know a huge amount about the marketplace. And there’s always been data available. I mean, we think of tools like Chartmetric as revolutionary tools, but they’re just evolutionary because in the past there wasn’t much data. There was always market data. There were always the charts. There was always ticket sales. And our people knew tons of shit, basically, meaning they knew what was going on. We call it “gut instinct”, but the dire of the gut is data.
JK: Given these large amounts of data to process for A&Rs, how does Chartmetric help with the artist discovery process? And with the decision-making involved in evaluating whether or not an artist is a good fit?
CJ: The challenge today is it’s just too much data. Twenty years ago, I, the average consumer, created two data points because I bought two products: a C.D. and a concert ticket. That was all the data which was generated. Today, the average consumer is creating 20,000 data points per year. To make it even more complex, there’s a lot more consumers than ever before because the music industry is monetizing far many more people than it ever did in the past; and it’s doing it on a global level. The music industry basically only operated in 40 countries in the world years ago. Today, it operates in 200, because we have consumers everywhere and they’re all interlinked. Today, A&R people just can’t absorb it all. They need additional tools to be able to make sense of all that sits in this information, so they can gain really simple insights.
There’s a huge amount of artists out there to discover. On chartmetric, we track about 3M artists at the moment. Therefore, the majority of those artists have never been discovered by anybody. There’s not really a problem finding good artists to discover; the difficulty, if you’re an A&R person is finding artists who could fit. You can then invest if you have the skills to add to what the artist is doing in order to lift something which is small. Ever since the emergence of digital, it’s not entirely coincidental that there’s been a trend in the music industry to discover or sign artists later and later. Record labels have been progressively signing artists who are more and more successful. There’s been a focus on just signing tracks as well: not taking the risk of signing artists, just taking a track which is successful and trying to provide investment to make it even more successful. If you’re investing in something, you want to generate a return. Therefore, generally, you want to find something very small and make it absolutely colossal. That’s high risk but will generate a significant return.
A lot of A&Rs over the past 20 years, have just become more and more conservative looking for things which were already successful and just tried to make them a little bit more successful. I think there’s been a gradual realization in the industry over the past five years that that’s not sustainable because there’s less and less of an incentive for an artist or a songwriter to actually sign once they’ve already become successful on their own. Why would you want to give away so much of your revenue? Is this organization genuinely going to provide additional support in order to make me even more successful?
There is a need to really look early, but also learn what characteristics there are among emerging artists, emerging songwriters, which provide the seed for an artist to become successful in the long term. We’ve become, as an industry, too reliant on looking for things which are already successful or quite successful. Looking for things which are quite successful is not necessarily the best way to find things which are going to be very successful in the future.
For the people who use Chartmetric, the key thing is not looking for artists who have achieved successes like getting a lot of streams or a lot of followers. Those things are too easy to fake and too easy to manipulate. A single outside influence like getting into a big playlist can have too profound an effect. Just because an artist gets into a playlist, does that mean they’re going to get into a big playlist in the future? Does that mean, if you sign the artist, that you’re going to be able to make them successful? Or it is because of chance that they achieve success this far? We’re advising to look for key triggers. That all predominantly comes down to evaluating the ability not only to acquire an audience but to retain an audience. Audience retention is the big challenge for the music industry going forward. The industry is great at audience acquisition, in the old days, it consisted of motivating fans to go to a record store, to hand over 15 $ and walk out with a CD. That was it, all you had to do, the job was done. Today, acquisition is just one part of a story, you need to retain the attention of the listener. Record labels can enable an artist to acquire more audiences, but they’re not so good at retaining them. Looking for artists who have that ability to retain the attention of audiences is a much more valuable use of A&R resources these days.
JK: Do you see that happening? Do you see in A&R departments focus shifting from raw follower numbers to audience retention ?
CJ: Yeah, very much so, particularly over the past few years. I think there’s been a big shift in terms of not just looking at performance on DSPs, not just looking at a higher number of monthly listeners or number of followers, but looking at much more complex data, looking at multiple datasets to try to identify why. If an artist is having success, why? Where is that success emanating from? Another trend, of course, is globalization. We published a piece of research last year called Trigger Cities. 20/30 years ago when I grew up as a kid, emerging artists did not release music. They played gigs week after week, month after month, year after year, in the vain hope that one day an A&R person would come along to their gig, see them, and think they were amazing, listen to that demo and then give them a chance to go into a studio and record something properly. Even then, when they released that, they would be releasing it in one country. Then maybe if it was successful, they got a chance to record another single. If that was successful, maybe an album. Generally, artists didn’t have the chance to see their music being sold internationally until they’d had two successful albums. These days, kids in their bedroom, instead of recording a demo, record properly and release their music globally within 24 hours. We’ve gone from a very, very localized marketplace to a completely globalized marketplace. Your addressable audience is much bigger. It turns out that, unsurprisingly, the way people engage with artists, share their love for artists is very different around the world. The behavior in the West is very different from many other countries.
JK: The music industry is structured by territory/countries. So how do managers and labels reconcile that global launch with their local investments?
CJ: Today, it’s very simple to get global data. All the data in Chartmetric is borderless, because consumers are borderless. On the other end, the music industry is still structured along very localized lines. Companies are local companies. Even a major record company operates as separate entities in every single country. That makes anything international very slow, very painstaking, even quite political in many instances; but data is available everywhere. I think although certain parts of the industry are struggling to adapt to this borderless marketplace, in general there is a quite rapid transition to looking at the music industry as one big global marketplace.
Ultimately, people find out about artists from friends, your biggest influences. I was influenced by my friends. My taste in music depended on my friends. But I only had four friends because I grew up a long time ago before the Internet, and my four friends lived on the same street.
Kids today have friends all over the world, so they usually don’t struggle too much with discovery.
JK: As you mentioned, A&Rs have tools plus their own network to help them with discovery, and that’s a lot of information to process. Some of them have a short list of 20/30 artists they’re going to look at during a day, to evaluate whether or not they could be a good fit. How do you think analytics tools are going to evolve moving forward? Because A&Rs are usually pretty overwhelmed and afraid to miss a gem.
CJ: It’s where data science really comes in, in order to evaluate metrics for huge numbers of artists across multiple different channels. Ultimately, there will always be a need for a human being involved in the process, because there are many critical metrics to A&R, which cannot convert into data. At the moment we are a long way from converging data. Computers are not good at listening to music nor hearing characteristics about music. In terms of really listening to music and discovering the emotional content of music, that’s actually still quite difficult. Computers can’t assess personal relationships very well. They can’t tell whether the manager is very good, whether the artist has an agent who genuinely believes in the artist, whether the artist is willing to do promo or wants to sort of like living a quiet life. These are difficult things to convert into algorithms at the moment. There is a critical need for human beings being involved in the process. There is so much data already available across streaming services. Comparison across individual artists, who will develop in unique and different ways, requires some real serious designs. That’s a huge part of what we do. Chartmetric is continuously analyzing trends across millions of artists in order to learn what happens and what has happened over the past five years.
JK: Data science is usually good at solving large scale problems.. Do you think tools like Chartmetric can actually tackle the artist scouting problem at a large scale? To be more specific, each A&R person has his.her own set of criterias to make their decisions depending on their market. To what extent do A&Rs need personalization on these tools?
CJ: When you are having to analyze data points across a dozen different services, building and reading the impact of radio, factoring in the impact of other media exposure, that then becomes very challenging for a human being to do when they’re looking at 30 artists a day. There is a need to use analytics to actually help the process in terms of personalization. Ultimately, we’ll head to the point where you can put every single channel and personalize it. Building tools now is about enabling every A&R to personalize their own search according to their own priorities.
One artist can be a perfect fit for one company, but a terrible fit for other companies. Every company has a very unique skill set. Very often in the past, artists have signed to the wrong record company. That’s always happened. That’s a mistake for the artists and it’s a mistake for the record company. Nobody wants to sign the wrong artists; but one record company overlooking an artist doesn’t mean that another record company will not see that same artist fitful.
JK: What’s your take on predictive tools? Is there such an algorithm that can predict success?
CJ: We’ve been building predictive tools for a couple of years now. You can always be predictive, but it depends on what people are looking for. Predictive tools can take a lot of the workload out and help identify new things which are not on your radar. But, if you’re a record company and you’re making a colossal investment in an artist, you’ve got to have a lot of trust in predictive tools. We want people to have trust in them, but you have to be able to provide a complete suite of supporting data as well for people who will make the decision. An A&R person won’t sign many artists in their career.
To be more specific about predictive tools, you also have a matter of time frame because the shorter time you’re trying to anticipate, the easier it gets. But if you’re looking at longer term predictions, you have to factor in a lot of things that are not right now in the tools or in the data set.
Every single person has their unique objective. It’s easy to develop a predictive predictive tool which will predict, but will it predict what people want to see? Everybody has slightly different objectives. Prediction is an obvious goal for all of this. What’s the next level of prediction? It’s a gradual, inevitable, but never ending process.
JK: In data science, there is something called the “cold start” problem. If you know nothing about an artist, you can’t predict what’s going to happen. So A&Rs would have to, like you said, bet on something that’s already there. There has to be some success already happening to be able to predict more.
CJ: Once you get your prediction to that level, that’s when the human element is threatened, but I think we are very far from that point. You have to factor in the team, the level of how are they willing to invest and expose themselves all the time? Do they want to tour? Do they want to lifestream? What is their take on their career? This information is not available in datasets.
JK: What about artists who fake their numbers?
CJ: There are always going to be numbers that are easy to fake, but it’s so easy to spot the numbers. A lot of A&R have several steps : discovery, initial analysis and testing – why do we see these numbers? Is there something on? Very often, you can see at an early stage that a number is not organic. Sometimes it can benign: something exploding on TikTok, getting into a big playlist, .. because we combine data, you can now see these things. If nothing explains, you can see that this person is buying followers. These aren’t new trends, in the old days people bought tickets for their own gigs. People sort of rented their crowds.
JK: How is Chartmetric trying to ease the A&R job?
CJ: Essentially, we want people not to walk away from Chartmetric with a ton of data in their head because they won’t be able to hold that much data. We want to weigh with just a single insight which enables making a really smart decision.
There’s one number in Chartmetric which people are increasingly measuring and that’s CPP, cross platform performance. We’re combining multiple metrics from multiple platforms because of course, when faced with a lot of data, people will often focus on the one number which suits their argument. With CPP, which is a cross platform index, bringing in data points from multiple different platforms eliminates that. You may still see a number which you really like, but then when you look at CPP, you say, hang on a minute, that number is not influential.
JK: You introduced this new KPI, CPP (Cross Platform Performance), in the music industry. How was your experience introducing that KPI and driving its adoption?
CJ: We haven’t really blasted it out there with a massive launch. We went for a gradual adoption of the metric. Eventually, we felt it was particularly used by marketing people as a way to gauge genuine performance across multiple platforms. We’re horribly siloed in our marketing approach in the music industry. We think success on one channel equates to success everywhere. And it doesn’t. Of course, you have to have success in multiple different places.
The great advantage of using CPP is because it’s looking continuously at three million artists across multiple different platforms and providing this in a single index, which is adaptive over time. It enables far easier comparison. Then you can actually dive in and look in more depth to really understand why an artist is developing.
JK: What advice would you give an artist to pass the “A&R” test?
CJ: Perseverance. There’s no such thing as overnight success. It’s about hard work. It’s about not giving up. These days, in particular, we always talk about 365 marketing. Marketing campaigns in the old days in the music industry were very short term. The music industry, compared to most other business sector, releases a huge amount of new products. It has fallen into this age-old strategy of marketing things for two weeks and moving on to the next thing. In an age where audience acquisition and retention are the critical factors, you’ve got to retain an audience. You have to be working 365 days a year on audience retention. If the artist isn’t doing that, then nobody else can do that.
Really successful artists have always worked 365 days a year. It’s critical for everybody involved in the process, not just the artist management or the label. It’s tougher if you approach things from an old mindset. If you approach things with a fresh mindset, if you look at the marketplace for how it exists today and don’t try and equate it to the old marketplace, then I think it’s a lot easier.
Keeping attention is challenging but cheap. There’s almost a sense of freedom to the marketplace. You know, in the old days, it didn’t matter how hard you worked unless you were able to cover the final mile and get your product in a store in front of consumers. However, if you didn’t have that covered, then it didn’t matter how hard you worked. You had to get to the front and center in record stores to get attention. It didn’t matter what else you did unless you were there. Once you were there, it was very difficult to keep that success because of so many other priorities coming from record labels which were going to displace you. These days it’s completely different. When I started in the record industry, they always used to be this adage which, if you went to a sales conference at a major label, one of the label heads would back in their fist on the table and say, it’s all about making it big. Your new release had to be as high in the charts as possible. You had to have the highest charts placing, because if you didn’t do that, then it would be impossible to get exposure. That may sound daunting, but the marketplace today provides a lot more freedom, a lot more flexibility, a lot more creativity in terms of marketing.