More than 26m smart speakers were shipped globally in 2017 according to research firm Futuresource, with the Amazon Echo and Google Home ranges proving popular devices.
That was the spur for an event hosted by British music-industry bodies the BPI and ERA last night in London, and a report exploring the impact for music, published by Music Ally in partnership with the two organisations.
You can read the report here: it’s part primer on the devices and market figures, and part analysis of what the growth of this hardware category means for musicians, fans and music rightsholders.
‘Audio hardware is back in fashion…’
The event began with Futuresource’s principal analyst, entertainment media, David Sidebottom, outlining some of his company’s recent findings on the market.
“Smart speakers are probably one of the most exciting consumer products we’ve seen in the music industry for quite some time,” he said.
“We always said if the music market was to go to the next level, it really had to embrace a truly seamless digital listening experience in the home. And we’re starting to see that now… Audio hardware is back in fashion. It’s a sexy product… And it’s no coincidence that we’re seeing the overall music market grow. The two go hand-in-hand.”
Sidebottom talked about the growth in music-streaming subscriptions in the UK, which have doubled from 4.5 million people at the end of 2015 to 9.1 million at the end of 2017, driven mainly by Spotify, Apple Music and Amazon.
Futuresource’s data indicates that there were 3.5m smart speakers in use at the end of 2017 in the UK, and the company expects more than 4m of these devices to be sold in 2018, taking the UK to an install base of 7.5m. “Consumer awareness has skyrocketed,” said Sidebottom.
What are people using their smart speakers for? 63% of UK and US owners are using them for music selection and streaming, according to Futuresource. That’s by far the most popular use case, ahead of the 32% who use these speakers for weather and news updates.
Encouragingly, 30% of smart speaker owners say they use their devices to find new music, and of those, 78% say they find new music every day.
“That’s a pretty impressive number. And 8% of UK smart-speaker owners say that finding new music is the one application they couldn’t live without for their smart speaker,” said Sidebottom.
“It’s clear that we are seeing exponential growth in music consumption because of this, and in particular music discovery – in an era when it’s quite easy to get bogged down by all the choices on offer.”
Smart-speaker owners are four times more likely to have a music subscription than non-owners, which as Sidebottom noted, offers two avenues of growth for the music industry: first, by increasing the penetration of music subscriptions among smart-speaker owners, and second, by selling more of those devices.
Finally, almost 30% of smart-speaker owners say they’ll buy another one in the next six months. “Multi-room usage is going to increase exponentially over the next few years,” said Sidebottom.
“By 2021 we’re looking at 20m smart speakers in use in the UK… We do feel this is almost a revolution in this market. It’s certainly a revolution in digital-music listening in the home.”
.@MusicAlly’s Editor-in-Chief Stuart Dredge presents key findings of the BPI & @ERALTD commissioned report: Everybody’s Talkin’ – The impact of smart speakers and virtual assistants on music consumption pic.twitter.com/jrRgUhKssw
— BPI (@bpi_music) March 20, 2018
‘Cutting-edge technology… marketed at a mainstream audience’
I was the next speaker, on behalf of Music Ally – yes, it’s a bit awkward writing up your own speech, but here are some of my key points (which you can find written up at more length in the report):
– Smart speakers in 2018 feels a bit like Apple’s original iPod in 2001, or its App Store’s launch in 2008. A new thing that’s hugely interesting, but whose implications for music we won’t completely understand until a few years down the line.
– One very important thing about smart speakers, particularly the Echo line, is that they’re cutting-edge technology (voice control and intelligent assistants) in devices that right from the start have been marketed at a mainstream audience. Not just tech geeks and/or hardcore music fans. And these are exactly the kind of people that the music industry wants to encourage to take up music subscriptions.
– These devices may fuel new listening patterns, with more general voice commands like ‘play me music’ or ‘play me party music’. Yet while these are quite general requests, the music served up in return can be hyper-personalised, based on our previous listening and everything else the speaker / voice-assistant knows about us. My ‘party music’ might be very different to yours.
– Smart speakers are a big challenge for labels on the metadata side of things. All of these new listening patterns – these voice queries and algorithmically-driven streams of music – are fuelled by metadata. Is a song happy or sad? Indie or hip-hop? Is it good for cooking to or for working out? Is it from the 1960s, the 1990s or last year? Is it by this songwriter or that songwriter? If labels aren’t providing that, then the tech companies will create it for them.
– There are some big creative opportunities for music marketing in a voice-controlled world. How do you market music when there’s no screen? It’s going to be really interesting to see how labels and artists create their own ‘skills’ (for Alexa) and ‘actions’ (for Google Assistant) – the equivalent of apps for these voice assistants. RCA UK is a real pioneer here: it soft-launched an Alexa skill for Paloma Faith in January called Paloma’s Bedtime. Designed for parents with babies and toddlers, it offers lullabies, stories, white noise and acapella tracks from Paloma’s latest album. It’s an early experiment, but there will be many more.
– I talked about the idea of smart speakers fuelling conversation-based discovery, cribbing shamelessly from an interview with then-CTO of Universal Music, Ty Roberts, in which he talked about Alexa asking you if you want to know more about the roots of hip-hop, then talking you through the key artists, songs and cultural/historical context around them. You could take this approach to genres and scenes – ska, grime, British heavy metal – or for individual artists.
– Shifting gear a bit, smart speakers are also a challenge for pureplay streaming services like Spotify and Deezer, since the main devices (in the west, at least) are made and controlled by Amazon, Google and Apple – who have their own services. This is why the pureplays are already lobbying in Europe for ‘fair access to platforms’ (a phrase that includes app stores and search engines, not just smart speakers). It also may be a factor in nudging Spotify towards launching its own smart speaker, which seems to be on the cards.
– Finally, smart speakers are shopping devices (the Echo particularly so) rather than just listening devices. There is potential for music fans to buy tickets, merchandise, vinyl and other products from the artists that they love. How far are we off my smart speaker saying ‘Stuart, I know you love the Chemical Brothers because you stream them lots: would you like a ticket to their October gig, and by the way, you can pre-order the vinyl version of their upcoming album…’?
‘What will drive the next evolution of streaming will be the home’
Next to speak was Amazon Music UK’s director Paul Firth, who talked about the Amazon Music Unlimited service as being “built with voice in mind from the very beginning”.
“We talk a lot about voice control being the future. We have to get used to it being in the present… this is something people are doing right now,” he continued.
Firth was clear about how Amazon sees voice’s role in the evolution of the music-streaming market.
“What undoubtedly has driven music-streaming to date is the smartphone… But that s-curve of adoption that happens with any new technology is starting to top out for smartphone,” he said. “What will drive the next evolution of streaming will be the home. And probably after the home, the car. And these are both environments where voice-control will come to the fore.”
Firth also talked about the mainstream appeal of music access through a smart speaker, as opposed to a computer and/or smartphone.
“For many people, really, music streaming has been too hard. I’m thinking about my mother-in-law or my father. A large proportion of the population… Streaming was not really made for them, and streaming was too hard. What we see is that voice control is part of how we change this,” he said.
He also scotched one commonly-held view about smart speakers: that the most common query from their owners is the super-general ‘play me music’. It is the single most common request, but for Alexa, if you add together all the different requests for ‘play [artist name]’ and ‘play [track name]’ they “dwarf” the ‘play music’ command.
“It’s not the commonest request by a long way, actually. People are picking what they play quite a lot,” said Firth.
Firth talked about the mainstream nature of Amazon’s music audience that is streaming via Echo speakers. “It really does change the whole customer profile who we’re thinking of when we think about a ‘music streamer’,” he said.
“For many people who would not have taken the step into music streaming before, this is the enabler that gets them there in the first place… it grows the audience that we have.”
Firth also returned to my part about music marketing in a voice-controlled world, flagging up Amazon’s ‘Side By Side’ program, which gets artists to talk about their new album, and then serves up that commentary alongside the tracks.
“We’ve got to experiment with this, getting new ideas out there about how we market music together,” he said, addressing labels. “What we see now is a very early-stage iteration of voice.”
‘This is not THE way people will listen to music. It’s one way’
The final section of the BPI/ERA event was a panel discussion, moderated by Music Ally CEO Paul Brindley.
The panel included Kara Mukerjee, head of digital at RCA Label Group UK; Pete Downton, deputy CEO at 7digital; Scott Cohen, founder and VP international at The Orchard; and Aaron Bogucki, VP of digital marketing at Kobalt Music.
The first question focused on smart speakers’ potential impact on the radio industry, with Downton noting that when 7digital ran some consumer research with AudienceNet in the summer of 2017, it found that around 15% of music listening was music-streaming services, but that more than 65% came from “lean-back” experiences like radio and YouTube.
“What I see is the radio industry going through the iTunes moment. Suddenly you unbundle the experience, and allow consumers in a frictionless way to listen to what they want, when they want it,” he said: citing an Echo owner listening to news from their preferred radio station around music from their streaming service.
He also talked about the opportunity for creatives in the radio world (of which 7digital has a number) to bring their skills to bear on content for smart speakers.
“It provides the industry with a unique opportunity, that we’re only just starting to see with excellent projects like Side By Side. The storytelling around the music,” said Downton.
“If you think about the skills that we bring to a program on 6music to tell stories about an artist or genre, and you think about delivering that into a streaming service, that’s pretty exciting for the music industry.”
Is there a risk that the music industry is overhyping smart speakers as the latest new technology? Mukerjee suggested that the excitement (but also the curiosity about the implications) is well-founded.
“I don’t think it can be overhyped. We’re fundamentally decoupling music consumption from every single element of it that’s tangible,” she said. “There is absolutely no experience of physicality or visibility, or even seeing the artist title or the name… This is really significant, I’d say.”
Bogucki agreed, and said that there’s an important shift for music marketers here: in the past, their job was to try to get people to listen to an artist or a track. Now, people may be hearing that track (served up by a playlist and/or voice assistant), but not knowing much about it.
“As music marketers, our job was to get you to that point where you consumed that song and discovered that song… Now we’re getting to the point where you’re at that bottom of the funnel already, but you don’t know anything about the artist: their story… why you should care deeply about them,” he said.
“Our challenge now is from that touchpoint to take you back to the top of the funnel, and take you through that journey… Our job as marketers is storytelling, and to get you to really care about that artist. I find that to be a massive challenge for us.”
Cohen injected a note of warning into the debate, reminding the industry not to assume that *all* listening will be through smart speakers and/or voice assistants in the future.
“We do get a little overexcited about new things. I love voice-activated speakers… but it’s also part of an ecosystem. This is not THE way people will listen to music. It’s one way, at one point of time,” he said.
“How many households in the UK have a toaster? And yes it’s impacted food preparation and food consumption. However, it’s not the only way we eat! So yes, we will listen through voice activation, but not on the tube [for example]. It will be one way we do it.”
Even so, the consensus was that while smart speakers can serve up a stream of music without context – who it’s by, what their story is – there is potential for these conversational interfaces to help listeners dig deeper.
“We have to make it frictionless to go deeper. It’s the industry’s challenge to tell stories that engage people in music beyond that initial level… We are just scratching the surface in term of music experiences, for the music fan or the mainstream consumer,” said Downton.
Mukerjee said there is a challenge for more niche artists, noting that the current generation of streaming services and voice assistants are “primarily geared towards playlist consumption and lean-back consumption”.
She said this is positive in one sense: it opens up opportunities to reach a mass audience. But niche-music consumption can be difficult: Mukerjee related the tale of asking for the new album by the Decemberists from an (uncomprehending) smart speaker, before giving up and asking for Spotify’s RapCaviar playlist instead. “Purely because I knew that invocation would work,” she said.
“It has huge implications for the equivalent of SEO for a voice interface.. you’ll see changes to artist names and track names purely because they have to respond to this,” she predicted.
‘We need to understand the data first, and understand what people are searching for’
This is where good metadata comes in, with Bogucki calling for the smart-speaker manufacturers to share more information with the music industry about how people are interacting with their devices, so that rightsholders can understand what new metadata they should be providing.
“It all starts with the data: getting us all to agree with what that metadata should be and look like is going to be fun,” he said, with a grimace.
“For me it’s the platforms, Amazon, Spotify, whoever’s there responding to that query of giving you that content. It’s up to them to hopefully open that up at least to the content providers. ‘Here’s what people are actually saying and doing to discover music’ and that’s going to help us understand that,” he continued.
“For us as marketers and content owners, to understand the context of what people are engaging with or asking for is going to be really important… We need to understand the data first, and understand what people are searching for. At this point we don’t really know, although we can guess it.”
Mukerjee agreed, saying that there are actually two metadata challenges here for labels. One is making sure music has the correct data on its genre, era, themes and potential listening contexts – from kitchens to cars.
“All that stuff’s in the works. But what needs to be developed is pronunciation-specific, vocal UI-specific metadata infrastructure,” she said.
“There are so many artists with dollar signs in their names, or different grammatical elements. Those artists cannot be found within the current smart-speaker environment! MØ, for example. There’s another girl band called M.O with the exact same pronunciation… There is currently no way to differentiate between these artists.”
Mukerjee suggested that dialogue may be the answer here: a conversation with the voice assistant about which artist a listener wants to hear, rather than the assistant having to make an assumption.
She also talked about the Paloma’s Bedtime Alexa skill, stressing that it was very much an experiment on what an artist ‘app’ might be like in a voice world.
“It was very cheap to build, really short development timelines. We turned the whole build around in about three weeks,” she said. “That is fundamentally where we are at the moment: we’re just learning about what can be done in this space,” she said.
Downton praised Amazon’s Alexa team for its openness to third-party innovation.
“They care about finding relevant responses to utterances to the device, and they care about the quality of that response right now. If that mindset continues, that drives a lot of innovation,” he said. “We’re seeing a ton of innovation. How that translates into products and experiences, we’re going to see in the next year or two.”
Cohen warned against expectations that every music artist should have their own smart-speaker skill.
“In 2008, it was ‘every band will have their own app!’ I think there’s going to be more general apps and skillsets across these. Not every artist will have their own thing. That just becomes silly,” he said.
“Now’s the chance, when the space is wide open, to experiment with it. But going forward in 10 years, I’m not going to have the Coldplay app on Alexa.”
Bogucki sees potential in other areas. “I don’t think we’re too far from celebrities and artists being the voice of your home voice-assistant. That seems like a massive market: for Liam Gallagher to be your voice assistant, for example,” he said.
“We used to have Snoop Dogg telling us how to drive our car! Mark Zuckerberg has Morgan Freeman as his voice assistant in his house. That is coming, and is something you’ll be able to buy.”
The panel discussion also touched on the question of Spotify and pureplays negotiating a device category controlled by their big-tech rivals.
“It’s a huge fear,” said Cohen. “I use Spotify on my Echo, and it’s listening to my Spotify playlists. Obviously it can listen to everything I’ve played and take a note of it. And if one day it wants to say ‘We know all your playlists and what you like, and we can offer it to you for the next six months for £14.99 because you’re a Prime user…”
[Caveat: as far as we know, the Echo doesn’t track every song that someone plays through a non-Amazon streaming service: it just records the voice commands used by owners]
Bogucki thinks that longtime Spotify subscribers will likely stick with that service, but noted that a lot of growth to come in the music-streaming market is from people who haven’t yet picked a service.
“Your average consumer who doesn’t have a subscription to Spotify, when they get an Echo: ‘Yeah I wanna stream music’. And it’s immediately in to Amazon. That is a really great driver, but I think it will put Amazon in the driver seat for much higher penetration,” he said.
Cohen brought the talk back to in-car voice assistants. “What really puts you in the driver’s seat is when this moves into the car. Talk about disruption. I think this might be the final nail in the coffin for the CD,” he said.
“Still today, it’s a better environment to listen to a CD in a car than to bring your phone and plug it in. You put an Echo speaker in your car? Game-changer. It’s a game-changer for radio, game-changer for CD. Game-changer.”
‘Personal assistants and voice control… this will change our lives’
The event ended with a Q&A segment, in which the panel were asked about the prospects for niche music: as one questioner noted, grime was once a niche, but is now mainstream. Can other underground music make a similar journey as smart speakers take hold?
“The way a lot of the streaming services developed is the answer to every question is a playlist. And we have approached some of that in the same way,” admitted Firth, before stressing that Amazon wants listeners to be able to dig deeper into the long tail of music as much as labels do.
“If you think for the long term, we need new music to come through as much as anybody else. That’s what we need to buy into… An echo chamber is as bad for us as it is for everybody,” he said.
“Longer term, personalisation is the answer. If we learn the signals of what makes someone respond to new music, we can take that forward… Think about how you talk to your best friend who knows more about music that you do. That’s what Alexa needs to be in the future.”
Mukerjee said hard work on metadata will pay off here. “All the organic metadata around those niche playlists and content needs to be improved… so all the rest of that content that isn’t being championed on a very superficial Home/Browse level, does get surfaced,” she said.
“There’s a lot of editorial, a lot of curation, a lot of metadata that needs to be done.. there will also be a lot more third-party suppliers of that data now. That’s going to be a startups space.”
Downton had the last word for the panel, on the impact of the technology that’s still relatively new for listeners and the industry alike.
“Personal assistants and voice control… this will change our lives. It’s the start of a change in behaviour that, in five years, we’ll look back on and say ‘Do you remember before that?’” he said.
“What does it mean for the music industry? We have a window of opportunity to change the game, to make it more engaging, to help discovery of music.”
But he also warned that if the industry doesn’t take this opportunity: “I do see a world where this can go racing past us while we’re still organising ourselves.”
Meanwhile, BPI boss Geoff Taylor delivered the event’s final remarks, bringing the conversation back to mainstream music listeners. “If we can get people like that engaged with music subscription, it’s a huge opportunity for the music industry,” he said.