Which three global hits best define the transformation of the record business in the last 18 months?
Well, there’s Lil Nas X’s Old Town Road, obviously – which built up such a head of steam independently on streaming services, Columbia Records couldn’t resist waving a multi-million dollar cheque to sign the artist in March last year.
Talking of Columbia and mega-cheques, there was also Roxanne by Arizona Zervas, which was doing over a million streams a day in the US before the Sony label swooped for its second viral hit-signing of 2019 – for another megabucks fee.
And then, of course, there was/is Dance Monkey by Tones and I. Having now surpassed a billion Spotify streams worldwide, the Australian-born phenomenon continues to rack up over four million streams a day on Daniel Ek’s service, after pretty much remaining No.1 on Spotify’s worldwide chart, consistently, for a full five months.
What unites all three of these tracks? Well, there’s a Sony Music connection – all three are signed to the major, with the exception of Dance Monkey outside of Australasia, where it’s signed to Elektra/ Parlophone/Warner. But they’re also all tracks of which, at some point or other over the past year, you’ve very possibly heard an A&R executive suggest that they “came out of nowhere”.
One person who strongly disagrees with this notion – that viral streaming smashes are born and gestated in a bewildering vacuum obscured from the industry’s gaze – is Conrad Withey, CEO and founder of London-based A&R scouting platform Instrumental. As Withey puts it, Lil Nas X, Tones and I and Arizona Zervas “all came from somewhere” – the industry just happened to be looking in the wrong direction at the time.
That somewhere, Withey points out with a grin, was closely monitored by Instrumental’s algorithm, which flagged all three breakout artists as ‘hot’ months, if not years, before major labels started piling on the zeros in their signing contracts.
(Instrumental’s tools flagged Lil Nas X as ‘hot’ in December 2018, three months before Columbia swooped; Tones and I was flagged in March 2019, two months before Dance Monkey was even officially released as a single in Australia – let alone achieved its world-conquering run; and Arizona Zervas, amazingly, was flagged as ‘hot’ by Instrumental in December 2017, nearly two full years before Columbia snapped up Roxanne.)
With over 40,000 tracks being uploaded to Spotify each day, and some 500 hours of video uploaded to YouTube every minute, Instrumental monitors an inordinate amount of music. To aid its discovery algorithms, the platform uses data from 30,000 influential Spotify playlists and hundreds of the ‘most engaged’ YouTube upload channels (GRM Daily, World Star Hip Hop etc.) to filter out the artists with the largest potential for making it big on digital services.
As a result of this fishing exercise, Instrumental finds around 5,000 new artists every week with the potential to go on to greater things. It can then match these artists to over 4,000 ‘micro genres’ in order to help labels, managers, agents and others to discover would-be hit artists operating in their specialist fields.
To aid those labels further, Instrumental recently launched its own ‘Watchlists’, which it manages internally on behalf of record companies. These Watchlists continually monitor data on artists that have been highlighted as being promising for a specific external A&R team, but when said A&R team deems it ‘too early’ to commit to these acts with a full signing.
Watchlists pull in additional social data about each artist, in addition to Instrumental’s own daily ratings – with one based on growth (Gx points) and one based on online engagement/popularity (Ix points). Any spikes in status automatically generate email and/or text alerts that can be sent to label A&R teams.
Here, Withey – a former President of Warner Music Entertainment – explains how Instrumental is helping record labels at home and abroad find and sign tomorrow’s superstars even earlier, he says, than a pair of ‘golden ears’ might…
How is Instrumental able to flag hot artists so early – what is the algorithm looking for?
Over the last two to three years of developing our tech platform we have finessed the discovery process to be able to pinpoint that moment when an artist starts to become meaningful to the wider music industry versus being a self releasing ‘enthusiast’. We have a range of criteria across social and streaming data for the initial discovery point but of course very few of those will become ‘hot’.
“Over the last two to three years of developing our tech platform we have finessed the discovery process to be able to pinpoint that moment when an artist starts to become meaningful to the wider music industry versus being a self releasing ‘enthusiast’”
What our algorithm is looking for in order to categorise an artist as hot is a set of growth requirements all happening at the same time. That covers streaming performance including playlisting and follower growth, but also social engagement (broader popularity driven by sharing, subscribing etc). We’ve found that when all these things happen together in a particular way that is a reliable indicator of emerging and ongoing commercial potential.
Then what matters is tracking whether artists stay hot. Tones and I, for example, was flagged as ‘hot’ on our platform just seven days after the release of her first single Johnny Run Away and when she only had 900 followers on Spotify. I doubt many people would have written a multi-million dollar cheque at that point.
But she has stayed ‘hot’ every week since then, which is exceptional – just as her resulting success has been. Watching her hot streak develop over the following few weeks you would have grown in confidence at signing a deal and that would still have been many weeks before she truly broke out internationally.
You recently launched Instrumental’s Watchlists. How do they work? How many artists are added to specific genre lists each week?
Watchlists are simply a response to what clients are telling us they need. The music industry splits into majors and everyone else. Major labels are perfectly equipped to extract value from data and insight platforms like ours, as they are so well resourced – both within each frontline label and centrally. Everyone else, though, is resource and time poor so our Watchlists offer a simple way to ensure you don’t miss out on new artists important to your business.
They are populated and run by our team and are typically genre- or country-focused – for example, UK hip-hop, Nashville or Australia Watchlists – and therefore targeted at a more specialist label or regionally focused music business.
We include only relevant emerging, independent artists and build a dedicated Watchlist that any number of subscribers can access. The data is refreshed every day and we rank the artists in the list by two metrics – Gx, our growth score which tracks momentum, and Ix, our Instrumental score, which measures popularity.
Any significant daily changes to an artist’s status triggers alerts to the subscribers of those lists, so even when you are super busy you won’t miss out on important news. Depending on the genre, we discover, qualify and add between 10- 50 new artists every week. This is such a dynamic business now – it’s incredible how quickly things change, every single day.
Which labels are you already working with and have you seen any notable success/case studies there in the past two years?
We are working with all three majors in markets across North and South America, Europe and Asia and a wide range of indie labels and non-record businesses.
I wouldn’t call out any particular case studies because I think there are so many factors that have to combine to deliver success and our [role] is super specific to the very earliest stage of that journey – but just knowing that we were able to flag up acts like Lil Nas X, Lil Tecca, Tones and I (pictured), Arizona Zervas and others, when they had less than 1,000 followers, backs up our confidence that the value is there.
And it isn’t just about mainstream pop acts. Those same dynamics can be seen when you drill down and look at specific genres like classical, jazz, Christian, ambient and so on – whatever level your business chooses to engage with new talent the data can help focus that process on prospects that will really drive growth.
A&R obviously gets a bundle of resources from the music business, but does scouting get enough of that pie?
Such a good question and something of a bugbear of mine. A&R covers a huge number of things – a wide variety of skills, processes, instinct, judgement and analysis. When you look at other industries these things have all become specialised, meaning that someone’s entire job is to become world class at each bit. So often in music one person is expected to do it all, which in the modern, streaming-first world just isn’t possible.
“The application of machine learning and AI is all about augmenting human processes and Instrumental is committed to being the best in the world at scouting for new artists and letting you get on with the rest of the job.”
So that’s where we come in – at least for one part of the A&R process. You can let the tech take the strain on the scouting and profiling side to support an individual or a team and be confident that our entire focus will be on doing that bit of the job really well.
As I have mentioned before on MBW, the application of machine learning and AI is all about augmenting human processes and Instrumental is committed to being the best in the world at scouting for new artists and letting you get on with the rest of the job.
You run your own artist services arm, frtyfve, which now boasts two UK Christmas No.1s [with LadBaby]. How does that work, how do you ensure it’s not competitive to your main clients, and how is it performing?
We actually started frtyfve as a way to generate case studies that we could use to demonstrate the value of a data-driven [Instrumental] scouting strategy. We offer artists our services on a single track basis, and use our platform every day to identify prospects. We are looking at super-early stage [signings].
We don’t take any long-term rights in artists. so can’t get in the way of any of [Instrumental’s] clients wanting to sign an act.
The services arm is going really well and we are expanding this year. And yes, two UK Christmas No.1s under our belt proves what’s possible. We’ve loved working with LadBaby and raising lots of money for the Trussell Trust, but those projects are something of an outlier versus the day to day on frtyfve.
That said, we wanted to work with Mark and Rox because their growth and engagement data was undeniable and we felt we could convert that into sales – which we did!
Don’t Instrumental’s tools take the humanism and romanticism out of the A&R process?
Imagine you are a label scout or A&R manager. Your job is to find trending tracks and talent from amongst the 40,000 new releases everyday. You have nothing to help you. So the only way to succeed in your job is to go through those manually. Or just ‘hit and hope’ you spot something that saves your job. That doesn’t sound very romantic. That’s the bit we help with.
We may be changing the discovery process, but everything after that is as romantic as ever. In fact, by leaving us to do the scouting you can get on with the romantic bit. The human-to-human, relationship-based, creative partnership. Go make sweet music, but do so with artists who’ve already proved themselves to some degree.
Do you still meet resistance in the industry based on those who dismiss you as being on the wrong side of ‘gut vs. data’?
All the time. My favourite email was from a metal label head in the US who said, ‘With respect, I like going to dirty bars, drinking beer and finding live bands – so you can fuck off.’
Fair enough, I love that attitude, but we are seeing that less and less. We have been described as ‘Moneyball For Music’ before in the FT and I do often compare the journey we are on to the way the use of data was adopted in sport. Take Opta, for example.
They are now a dominant force in football and their xG metric which measures a forward’s underlying performance and is used across the board – from Premier League clubs, their management, coaches and scouts, to TV pundits on Match of The Day.
But it took a long time to get there; Opta started up in 1996 and xG was first mentioned in analysis on MOTD in 2017. We know we have to be in this for the long run, because people take time to adjust their ways of working. But it will happen.
What did you learn during your time at Warner that has proven most valuable as you build out Instrumental?
When I was at Warner I ran Warner Music Entertainment and our whole remit was counter programming to the other frontline labels [Atlantic, Parlophone and WBR]. We looked for artists or music opportunities that we could develop that didn’t need radio to break and deliver a return on investment.
That typically meant I was drawn to signings where the creator already had a fanbase in place or we were targeting an audience that was under-served.
“I’ve always hated doing stuff just because other people are having success with something – I’m much happier swimming against the tide.”
That led to successful projects like Hugh Laurie’s blues albums, or The Overtones. No-one thought [such projects] were going to work, but I could see an engaged audience ready to consume new music.
I guess it was that mindset that led me to be fascinated first with YouTube and then other platforms like Spotify and TikTok, where audiences gather, create, discover and share – which leads you to opportunities you’d never thought about. I’ve always hated doing stuff just because other people are having success with something – I’m much happier swimming against the tide.
Instrumental raised $4m in growth capital in 2018. What has that money mainly been invested in?
It’s pretty much all been about investing in data, data science and engineering. We’ve bet on this being a long-term shift in the music industry and we continue to develop products and services that can help other businesses navigate the increasingly complex environment.
Our focus in 2020 is on developing more and more robust metrics and prediction capabilities – inspired by the likes of Opta – and looking to reshape how we measure success in this industry.
It’s interesting to see you launch ‘at a glance’ points via Ix and Gx. What’s the purpose/ambition of these metrics?
It’s very much in line with our goal of getting clients to the information that matters to them today. These indexes help point out meaningful change you should know about – whether that’s for an artist who is completely new to you or for an artist prospect you’ve been keeping an eye on for some time.
We are serving up notifications driven by our Ix and Gx scores via mobile texts and will launch an app this year because this is now a 24/7 proposition. Rarely are clients actually sat at their desktops; it’s a mobile- first world.
The independent artist sector is booming according to recent stats. Is that something you’re recognising – and what can the major labels do about it?
It really is and I think what is amazing is the speed with which an independent artist can be catapulted from obscurity to the mainstream nowadays.
It’s almost happening in hours – let alone days or weeks. And the scale is incredible. We find around 5,000 new quality artists a week from the self- releasing space.
“We find around 5,000 new quality artists a week from the self- releasing space.”
I do think there will be some interesting developments in how majors, indie labels and other music businesses work with this generation of talent, because they are completely different, culturally, to what came before.
They are savvy about their own potential, rights and commercial worth and, crucially, they’ve often done the heavy lifting before they even start conversations with partners.
That changes the dynamic completely. You really have to come with your ‘A’ game and demonstrate clearly the value you can add.
