This week has seen a Twitter storm erupt around whether Spotify pays off for new artists, fuelled by Atoms for Peace’s decision to pull their albums from the streaming music service in protest.
We suggested yesterday that Spotify needs to devote more attention to answering the specific allegations that streaming as a business model only works for big artists with sizeable back catalogues.
But wider questions about streaming’s value to the music industry as a whole haven’t gone away: particularly the question of whether legal streaming services cannibalise piracy more or less than they eat into sales of music downloads.
That’s the context for a report published by Spotify today called Adventures in the Netherlands: Spotify, Piracy and the new Dutch Experience.
It’s the work of the company’s director of economics Will Page, working with data from Musicmetric and GfK Netherlands, as well as historic studies on filesharing in the Netherlands – a country where Spotify launched in May 2010.
Page tells Music Ally the report has been seven months in the making, so it’s not a hacked-together study trying to rebut the arguments of Nigel Godrich, Thom Yorke and other musicians who’ve been criticising Spotify this week. So it doesn’t address those criticisms, but it does merit serious consideration in the light of the wider debate around piracy, streaming and sales.
Why Holland? Partly because Dutch digital recorded-music revenues rose by 66% in 2012 according to the IFPI: the highest growth in Western Europe. Partly spurred by a previous (independent) study published in October 2012 on filesharing in the Netherlands, which claimed music filesharing has been falling since 2008, while film and TV piracy had increased.
And yes, partly through an unsurprising desire to establish whether piracy has fallen there, and if so, whether Spotify can claim some of the credit.
The fact that the Netherlands isn’t in Scandinavia was important too, says Page, referring to concerns within the music industry that the startlingly-positive music-revenues growth in Sweden and Norway in recent times were a local phenomenon, unrepeatable elsewhere in the world.
“Spotify looks to have cracked a country outside of Scandinavia, and that’s huge,” he says, while stressing the efforts taken to ensure the study is a serious piece of research. “Obviously, I’m working for Spotify and this is a great news story to tell, but if people think this is filed under ‘PR’, that would make me want to resign and walk the dog. So many steps have been taken here to make it impartial. We’ve been uber-cautious that this is impartial and not political. The facts, not a distortion of the facts.”
Among the report’s headline claims: music piracy took place in 1.8m households in Netherlands in 2012, which is a quarter of the country’s total.
Page points to a study by Forrester in 2008 that claimed there were 5m Dutch music pirates, and data from IVIR indicating that this fell to 3m by 2011-12.
The 1.8m figure is for households, not individuals as in those past surveys: Page admits that this may mean more individual pirates in 2012 where two or more people are filesharing in the same house, but also suggests that people filesharing from home and work would be counted twice.
Spotify’s report – based on analysis of Musicmetric’s data on BitTorrent activity – claims that of the 1.8m piracy households in 2012, 532k of them only downloaded one music file, while 188k of them downloaded 16 files or more.
As the report puts it: “This Long Tail distribution is an important insight, as it highlights that most people take very little. Meanwhile, the top 10% take over half the content.”
The question raised by the report – Page, as an economist, prefers to see himself as “asking a better question, not championing a particular agenda” – is what this means for music rightsholders thinking about hardcore pirates (who are likely to keep pirating) versus casual pirates (who may be more easily tempted onto a legal streaming service).
The report is also clear in its aim not to hoover up all the credit for any filesharing fall in the Netherlands. In conversation, Page talks about a “Big Bang effect” of legal services launching, deals with telcos to promote them, AND legal action against piracy. Carrots and sticks.
Dutch ISPs were ordered to block The Pirate Bay in May 2012, for example – and the impact of the ensuing press coverage may have driven hardcore pirates to find ways around the block, but may also have driven casual pirates towards legal streaming services.
“It merits further research, because we don’t know enough about it,” says Page. “In Sweden, the big bang effect was Spotify, the Pirate Bay trial and a successful telco deal with Telia Sonera all happening at once. I don’t know how effective all three were individually, but happening together may have made them more effective. In Netherlands, the ordering is perhaps different, but they may well have been contributing factors.”
Spotify’s report doesn’t just dig into piracy as a general trend in the Netherlands. It also explores the relationship between sales, streams and torrents, to provide some answers (or at least pose some new questions) about cannibalisation between those three ways of getting music.
Page took single and album sales from GfK for four big pop albums released around the same time – by One Direction, Robbie Williams, Rihanna and Taylor Swift – then compared them to torrent downloads tracked by Musicmetric. The first two were available on Spotify from the same day as they went on sale, while the latter two were withheld from Spotify.
The workings are shown in full in the report, but the conclusion, in short, is that One Direction and Robbie Williams’ ratio of sales to torrents were much more positive than Rihanna or Taylor Swift’s. “One Direction and Robbie Williams sold 4 copies for each BitTorrent download whereas Rihanna and Taylor Swift sold only 1,” as the report puts it.
“There is no evidence in the 4 case studies showing that streaming on Spotify hurts sales… The most popular album on Spotify had the highest (best) sales to piracy ratio. One Direction’s Take Me Home was available on Spotify on its release day; it had the highest weekly Spotify stream count and sold the second largest volume of albums in its release week.”
It’s important to read the workings in full, and we hope that labels, artists and analysts will be doing just that, posing questions and picking holes to ensure that the methodology holds up. Because if it does – and Page says that he showed his workings to a roll-call of industry “arse-kickers, not cheerleaders” before publication – it’s significant.
“A lot of the debates going on about streaming services are a little bit half-filled. Look at people using the latest US sales figures to compare the effect of streaming on downloads, when those sales figures don’t include the numbers for streaming,” he says.
“The real problem is how you construct these debates within the music industry. With holdouts, it’s not just ‘what will streams do to my sales?’ – you have to ask what does it do to the torrent market as well. We think we’ve cracked a really good framework for debating the holdout issue through sales, streams and torrents. For One Direction, the facts are that they went hard on Spotify, they sold lots, streamed lots, and kept the piracy figures down.”
There’s more in the report, including a comparison between the Netherlands and Italy, where Spotify launched in February 2013 and intends to conduct similar research over the coming months and years.
Here’s something we think is positive about the report’s publication: this is the kind of data traditionally shared behind closed doors with music rightsholders by Spotify, rather than publicly with artists and songwriters. Opening it up so that artists can digest, discuss AND criticise the data and conclusions is a welcome move.
It’s also fair to note that we don’t think the report has many answers to the debate that’s raging this week about how (or whether) streaming music can benefit emerging artists well below the scale of One Direction, Rihanna, Taylor Swift and Robbie Williams.
That’s not a failing in the report itself, which has been in the works for a long time to answer a different question: the piracy/sales/streams relationship. That issue is relevant to all artists, not just the famous names. It’s just that it won’t dampen down the specific criticisms in the news this week. Perhaps that’s the task for Spotify’s next piece of research.
There will be a fair amount of people justifiably sceptical of a report produced by Spotify that claims to prove what Spotify has always said is its raison d’etre – eating into piracy. Yet we hope they read it, think about it and publish their responses to continue the debate. 2013 feels like an important year for more transparency in the streaming music space, and this report is a welcome step along that road.