This article from is Partner Content.

The Music Ally Focus Podcast recently welcomed Michael Horan, VP of Product at, which is just about to hit a remarkable milestone: 20 years as a music tech company. In that time, has gathered an enormous amount of data on how people listen to music – and we wanted to talk to Michael about what has learned from a music business that has changed dramatically in two decades. has grown from a company that tracked what you were listening to in your mp3 and CD collection to what you listen to everywhere online, and Michael explained to Music Ally’s Joe Sparrow the changes in how data has been used, and what changes are needed so that data can continue to be used in innovative ways to build communities and connect audiences with artists.

You can listen to the podcast in the player below, and we have also pulled out the key discussion points for you to read.

20 years is a long time!

Joe Sparrow: First of all, let’s acknowledge that is 20 years old. That’s a big anniversary in in any terms, but in particularly in music technology that stretch of time. As you look back at two decades of the platform, what do you think are the big things that last FM has brought to the music business in that time? Whether that’s innovations or technologies, or just a mindset, what is it?

Michael Horan: It is impressive – 20 years seems to be pretty incredible. You know, when we started out, frankly, the reason that we did start out was to keep track of what people were trying to discover on the available platforms at the time. This predates iTunes and streaming services in general were pretty small.

They didn’t have that widespread adoption, so being able to track what people are listening to from these disparate sources was really important. You wouldn’t be able to remember what you would listen to. And there was such a huge variety. People went from having small collections at home of CDs or vinyl to the internet – full of all the music, ever.

So that’s really how we started. And you know, that kernel still exists in us. Our primary function is to keep track of people’s listening histories across all available platforms, and to keep it in one dedicated repository. So the industry has changed, as you say. There’s been so many services that have popped up and, and faded away.

And for us, the, the one consistent thing is that we’re still here – and we’re still doing what we have always wanted to do, which is to allow people to keep track of everything they listen to. We started as well as a radio service, which we no longer do. And that was arguably a good decision on our behalf because we were the first global service – and we introduced this idea of a lean-back listening experience where you could passively choose your favourite artist, and immediately, a never-ending stream of music that was related to that was played. And the more you listen, the better the listening experience became. We’ve seen people who listen to music are no longer just pleased with this lean-back experience. They want to be more engaged and more involved in how the music is playing. 

So it’s important to remember that when we started, none of the current climate existed – and we were pretty much one of a handful of services that offered music to people that weren’t tied to a specific catalogue. It was as wide as the imagination could be. It used to be that you were told what to listen to, and now it’s become a place where you decide what you’re going to listen to. And that’s a really big distinction in the way that listening has changed over the past 20 years.

What do users want from their data?

JS: The service you’ve described is almost taken for granted now, which is: we track your listening history. I think when most people think of that now, they think of it in one particular platform, like they consume music over here and it’s all recorded somewhere.

People still want what does – so why do they want that? What, by having all their listing history grouped together, does it present to them that maybe they’re not aware of?

MH: So, before 2002, when we first launched, people would listen in in many disparate ways. They’d listen through their CD collections, and CD collections of friends. They’d listen on the radio, they’d go to a local record shop and hear new music that way.

It was a very sort of transactional experience where you were given a very small selection and you had to choose without really being able to know what you were hearing. You had to choose what you would invest your money in. Obviously around the turn of the millennium things changed a bit with peer-to-peer and, you know, that caused havoc in the music industry at large because suddenly, people were stealing music, but also people were given limitless access. allowed you to keep track of that limitless access, and effectively reduced the catalog from “everything” to “things that were interesting to you.”

As streaming services come and go, we’re able to go with you. We become somewhat like a musical passport. So it isn’t just this assumption that everybody listens on a single platform and they only listen that way, and that streaming service has all the information they need. In reality you’re going to other places – similar to going to a friend’s house or going to a record store or going to a gig.

People don’t listen just one way and they want to keep track no matter where they are. So we’re able to do that, and for 20 years we’ve managed to follow people between services that no longer exist and services that you know.

There’s this infinite amount of choice that sometimes they can’t cut through that noise. They’re stuck in whatever realm they’re being moved into. Whereas with us, we follow them. We don’t tell them where to go. We just follow them along. So it’s an incredibly passive thing, scrobbling, it’s very passive. And when you then come back and see what you’ve done, and reflect on what’s happened, you have a better sense of where you’re going to go next.

How is data used differently now? And how will it be used?

JS: You’ve been scrobbling data for decades – plural – now and that’s a lot of data… a frightening amount of data in some ways! What have you noticed about how that data is used – perhaps from the listener’s perspective, but also by the people and businesses that plug into that data, and try and find things in it? How is that changing?

MH: There’s personal user listening data and there’s that aggregated global data – and of course you can segment it any way you want, but a lot of our focus historically has been on the user’s personal data. The personal data is everything for us – the global data is a byproduct, and it’s incredibly valuable. 

We’ve seen a lot of services and companies that have grown over the past 20 years, and a lot of them have been acquired by, music services to help pad their offer, or to give more insights, to, say, artists or labels, in order to give them a better sense of where their music’s being consumed.

That’s all great, but by focusing on the user’s listening history – and giving them the insights that they need – they have constant access to their entire listening histories. And that distinction’s really important because we are seeing that more and more music listeners are interested in what data they’re offering up – and they want control over how that data’s used.

What we’re trying to do is give them the tools to reflect and see what their trends are, see what things have changed in their lives that have affected their listening histories. 

We were initially one of the few sources where people could get third party music listening data, and a lot of the time right now, there’s these weird distinctions where it’s considered proprietary data: nobody else can have it. That’s definitely a business decision that people will make – and I understand that value. We’re trying to unlock that in a certain respect because we don’t feel like we can shield our users from accessing their own data.

So it’s really important for us that there’s an openness to what we’re collecting and how we’re providing it. 

JS: Music listening maybe is subconscious, to an extent: users just want to listen to this or that, and then it’s recorded – and it can have consequences in the future if you wish to plug that data into something else. What are the potential future uses for this kind of data that we haven’t imagined yet? We’re moving perhaps into a music industry where the use of music is going to be in virtual experiences, in gaming, it’ll be interactive, or AI generated, and it’ll be web3 music and decentralized. So where are the opportunities for that listening history to be used in these new places?

MH: You know, you can’t necessarily differentiate a pop fan, from a metal fan, from a jazz fan. They’re all personal experiences to those users, and a lot of what our role should be is allowing people to connect, based on those unique passions. So, in the future, it, it really does come down to what technologies develop for how people are able to explore and discover new music.

We want to be able to facilitate those relationships between people. There have been dozens of marriages that we’ve been told about off the back of people using! That’s how personal these passions are. If you find somebody else who has an equally passionate response to an artist, that’s a deep and meaningful connection that we can facilitate.

So I think a lot of what our role is going to be in the future is, how do we identify these communities, around people’s personal experiences reflecting their musical tastes.  

I feel for musical artists right now because they’re spending more time being owners of their social profiles rather than creating music and anything that we can do to help these passions – whether it’s with artists or other musical fans – that’s a really important aspect. I think right now that is being lost because so much effort’s being put on, “how many pieces of content am I producing a day, a week, a month?” from an artist. 

I think’s role in the future will have to be how do you maintain  this log of what people are listening to, and how that can combine people together and drive organic growth of communities. 

JS: Community building is obviously a huge focus of artists now and the fans: it’s something that both ends want – a direct connection between the two. There’s a reason that artists are setting up Patreons and Discord servers and building these direct connections. Do you think that fans or music listeners are becoming more conscious of what they do, in relation to their favourite artists?

MH: Yes. There’s a certain level of “prove it” – where “I’m the biggest fan of X”. So with I can prove that – and you see a lot of people getting into these arguments! “You don’t know anything – you didn’t start listening to this person until last year!” We have proof of how big of a fan somebody is; we have proof of who discovered what first.

I don’t want to say that we want to encourage that competitiveness, but at the same time there is a certain level of authenticity in our data where you can’t really cheat it… there’s evidence to support whether or not you are a fan of an artist.

JS: It’s interesting you say that because so many new technologies are emerging around proof of fandom. I hate to bring NFTs into the conversation, but one suggested desirable use of NFTs is to say, okay, I was at Ed Sheeran’s second show: I am a real fan. But there’s also people are hanging around the waiting room before a launch of a new YouTube video. And you know it would be interesting to see who really watched the video when it was first launched. And that has meaning in fan communities, doesn’t it? “I was there. I was part of it.”

MH: Yes. And that’s what we see people using our data for. You know, there’s a lot of plugins right now that are being used across a variety of digital platforms: I actually did this – and we are the evidence. 

What data does the modern industry need?

JS: Let’s look at the present in 2022. The music industry is a data industry and a tech industry as much as anything else. Now, you’re able to look back at 20 years of data and have that context. What sort of data does the modern music industry need? Are there any changes that could be made around data collection analysis that would be useful?

MH: Yes, massively. Right now, everything is limited by the quality of the data that goes in. There’s a known quantity of music that’s out there, but every day there’s thousands of new releases. And a lot of this, you know, is provided with some level of laziness, I’ll say! You know, a remaster, a reissue: it’s just a means to pop it up in people’s periphery. 

I think the big problem right now is ambiguous artist names. There is no clean way to verify that a particular artist is that particular artist. You can have 13 Nirvanas if you wanted to, – there isn’t really a international code of standards. I know there’s a lot of initiatives and people doing a lot of work to resolve this fundamental issue: and it’s still in progress.

We would love to say that we have plans to be the best source of metadata going forward, but right now we’re really limited by what we’re receiving. So it’d be great if everybody got together and agrees that this is the standard for what makes an artist an artist, what makes a songwriter a songwriter, and just keep people honest. It sort of feels like it should be simple, but it also is difficult getting everyone in a room and saying, let’s sit down and figure this out, and let’s all agree.

I mean, there are tens of millions, if not more, of artists… they’re seemingly infinite!

One strength of is if you were ever in a band 15 years ago, the likelihood is we have information on them. It’s interesting, and that’s, I guess, going back to being able to say “prove it.” We had somebody start at recently and he was shocked to learn that we have his albums’ metadata on the site, and  people could theoretically discover him. 

That’s the powerful asset. But the problem is if he had a shared name with a different artist, we wouldn’t necessarily be able to resolve that. So part of the strength is in that infinite amount of data that’s out there, but the hardest part is to clean it up and consolidate it in a meaningful way. And nobody has corrected that yet.

JS: How confident do you feel that it might be fixed?

MH: It will – it has to. A lot of conversations, especially in the past five years, go back to royalties – and it’s critical to resolve that. There’s a lot of people who are being sidelined because they can’t prove that their music is theirs.

Users’ needs in 2022 – and escaping the algorithm

JS: What have you learned about the needs of users? When you speak to users of, let’s say, streaming platforms, they sort of assume that the algorithm is sort of “taking care of them” and it’ll present the music they need. But I also wonder if there’s a growing idea that people also want to discover for themselves and be a bit of a micro-influencer in their friend group. And how do you do that if you’re relying on an algorithm? So what are very modern consumers looking for in their experience and how does data fit into that? 

MH: That’s a really good question. I have recently seen a lot of articles suggesting that people are dropping their subscriptions and going back to their local catalogues they have stored… that “search bar paralysis” where people just don’t what to do.

Algorithms are very helpful, but they’re not everything. The most powerful recommendation is a personal one. And that can’t be lost. So that’s that sense of peer to peer –  I don’t mean in the technology sense – I mean people talking about their musical tastes.

Really what we focus on is what you’ve listened to, and creating safe places for people to be able to communicate to one another – because that’s the strongest connection that people will have with music. There is a general distrust for algorithms. They’re very helpful, they’re very useful as a starting place, but they’re not everything.

You’re going to find that a lot of algorithms miss a lot. I’m confident that ours is, is fairly strong, but at the end of the day, it’s more important for people to connect on a platform, than to just trust in the algorithm. So the data that’s missing is the user input and that’s part and parcel of our algorithm: we pay attention to when people aren’t interested in something.

Just because I like a specific album by an artist doesn’t mean I’m going to like all of [their music]. There’s an evolution when it comes to music, and it’s important to distinguish a song from an album from an artist. Those things aren’t necessarily all connected. It’s a great place to begin. But if I’m not interested in this other album, don’t tell me about it later.

JS: What we’re talking about here repeatedly is community and, and as you say, fans grouping around an album or a cluster of albums or a micro scene. It’s one of the things that a lot of the big incumbent streaming platforms, for instance, have shied away from. They don’t want to introduce that into the mix on their platforms or apps, but it definitely feels like something that is desired by fans and artists. And is that something which is gonna be fulfilled in the, in the next few years? 

MH: I would love to say yes. It’s very important. Of course, it’s a moderation nightmare! But I do think there are ways that you can approach it where you’re not alienating specific users and you’re not welcoming other types of users. For the most part, people have shied away from it because they’re more focused on, say, selling more music.  

I’m not going to say that we have all the tools that we want, or that we strive for, but it’s going to be a focus for us going into the future – where we see these communities are being formed on that we didn’t necessarily expect. And it’s obvious that there is a need in the market for people to fill that.

The next 20 years

JS: Let’s look ahead to the next 20 years. What happens next for

MH: I think it’s going to be really important that we keep track of where the music industry is going in general. Web3, the metaverse, the blockchain, all of these sort of buzzwords – we’re taking a keen interest in them and we want to ensure that we’re where people are consuming music, and tracking what they’re listening to. 

There are a lot of other threads that we could pull at, but I think, you know, the most important thing for us is to be where users are and allow them to just continue aggregating all of their plays, all of their passions in one place, and see what we can build on top of that.

EarPods and phone

Tools: platforms to help you reach new audiences

Tools: Kaiber

In the year or so since its launch, AI startup Kaiber has been making waves,…

Read all Tools >>

Leave a comment

Your email address will not be published. Required fields are marked *