Music popular-ish on YouTube

MusicAlly has done some analysis of what's popular on YouTube:

The headline stat is that music accounted for 62 of the top 250 channels ranked by annual views in 2014, with 43 of those falling under the Vevo banner.

In other words, music doesn't dominate YouTube, but it is mostly significant, but what really thrives are channels run by the big labels.

This is quite interesting:

Also notable: the low ranking for some of the traditional stars of social/digital media: Justin Bieber had the 43rd most popular YouTube channel in 2014 – admittedly a quiet year for him musically – while Lady Gaga was 158th and Britney Spears 203rd.

Although I think that says more about what MusicAlly thinks is happening on social media (All three stars are long past their buzz stage, surely?)

Late starting doesn't work in the pop industry

The i has a piece today headlined:

Women: If you are over 26, you'll probably never make it as a pop star

This is based on a data analysis of 'age at time of first number one single', and - to be honest - if you strip out Carly Rae Jepsen, who is something of an outlier, you're looking at 23 being the latest to get your career started.

But... surely the idea that the pop industry is a harsh place to late starters doesn't require a graph? Or one to show that the levels of investment in older women in the entertainment industry scuppers the chance of a launch (and we're calling 27 'older' here which shows the extent of the problem)?

In fact, given there can only be at most 520 number one artists in a ten year period, and generously, only 100 of those are liable to be debut number ones, and the population of the three countries which supply most US number ones (US, UK & Canada) is about 421 million, the headline should be:

People, you'll probably never make it as a pop star

Nielsen will count plugging your iPhone in as an iTunes visit

Back in March, writing about the IFPI's response to the EU study finding that unlicenced filesharing might lead to a small rise in sales, I asked if it could possibly be true that Nielsen counted plugging in an iPhone to recharge as a visit to iTunes.

The IFPI were polite, but fundamentally unhelpful when I asked them about what they were basing this claim on; happily, Nielsen were a bit more forthcoming and confirmed that iTunes data does include plugging in the phone:

After some internal research, I can confirm that IFPI's description of the NetView methodology is correct. How Nielsen looks at iTunes: anytime the app is active in focus, which can include any iTunes activity from listening to music to purchasing movies to synchronizing devices etc, which would include purchasing music but not limited to this activity.

Blimey.

I don't think this is quite the game over for the EU study that the IFPI thinks it is, although it weakens the strength of the research. The study had expressly ignored Amazon data because it was impossible to tell 'purchasing music' from the other activities one could do on the site; it might have been wise to exclude iTunes for the same reason.

However, the survey did find similar results for other services - it wasn't relying solely on iTunes data - so while it's not quite so solid as it appeared at first, it's still valuable if approached with caution.

Another survey suggests filesharers are bigger music purchasers

There's some fascinating data tucked into the Columbia University study of music collections in Germany and the US.

The first, and most important thing to notice, is that people who get a lot of files they shouldn't have tend to have more files that they have paid for.

(This isn't a surprise; we've seen these reports before.)

This, though, is the bit that should make sphincters tighten in offices where music is sold:

Around 79% of 18-29 year olds have music files, compared to only 14% of those over 64.

That, then, means that one in five young Americans have no music collection at all - or at least, none that are digitised.

It could mean that the survey spoke to a surprising number of hipsters who only buy hand-woven vinyl records; or possibly that a lot of American youngsters have become unshackled listeners, happy to just suck down tracks from a streaming service like Spotify whenever they need music.

Or it could just mean that twenty per cent of American young people have decided they don't need to collect music.

That presents one hell of a structural problem if you business relies on people amassing collections of songs.

Digital music worth more than pub and club jukeboxes

A small step towards digital music sustainability, as PRS announces that the royalties from digital streams is now worth more than royalties from pubs and clubs.

This might be a sign that digital music has become an increasingly robust business.

Or might be a reflection that there are 4,500 fewer pubs in business than 2008.

Piracy actually not that bad, but let's say it is

There's some interesting figures in the IFPI report into digital music, reported by Music Week. You remember piracy, don't you?

According to Nielsen/IFPI figures from November 2011, the percentage of internet users accessing at least one unlicensed site monthly stood at 27% in Europe and 28% worldwide – with around half using peer-to-peer networks.

That doesn't really sound very much, does it? Over 70% of internet users not even visiting an "unlicensed site" (whatever that is) at all.

But that's averages. What's happening in piracy hotspots?

In some countries, the rate of usage of illegal sites is far higher than the global average - for example 42 per cent and 44 per cent respectively in the major markets of Spain and Brazil (Nielsen/IFPI).

So, even where it's bad, fewer than half the people online visit a site which holds unlicensed music.

And America? How are things in America - a country where piracy is apparently so rife, we came within an ace of having SOPA destroy the free internet?

NPD data included in the IFPI report shows that 16% of the US internet population were using infringing P2P sites in Q4 2007, down to just 9% in Q4 2010.

Meanwhile, the average number of tracks downloaded from P2P services dropped from 35 to 18 in the same period.

All this, surely, is brilliant news for the entertainment industry, right?

Music Week isn't so sure:

However, levels of global music piracy won’t bring comfort to labels.

Really? They're relatively small, and appear to be declining. Why would they not take comfort from that?

Let's nip over to the IFPI site, where Frances Moore, CEO, is looking at the figures. Is he happy?

As we enter 2012, there are good reasons for optimism in the world of digital music. Legal services with expanding audiences have reached across the globe and consumer choice has been revolutionised.

Well, that's great news. Treble clefs all ro... oh, hang on. He's looking cross. Why is he looking cross?

"Any complacency now, however, would be a great mistake. Our digital business is progressing in spite of the environment in which it operates, not because of it. In 2012 the momentum needs to build further. We need legislation from governments with coordinated measures that deal with piracy effectively and in all its forms. We also need more cooperation from online intermediaries such as search engines and advertisers to support the legal digital music business."

Moore is, you'll spot, telling a whopper here. His organisation's own figures show that even with their gross over-estimation of piracy, it's in decline - mainly despite, not because of, the work of the labels and their chums in the movie industry.

(Actually, should the labels be worried? Even without any comeback, even when it's available for free, people are scooping up less music.)

To the untrained eye, it might look like the RIAA and their client organisations have bought themselves a fire engine; now there's no fire, rather than getting rid of the uniforms and hoses, they're running round going 'I think I still smell burning'.

They passed from being entertainment companies into copyright farmers long ago; now, it looks like they're going to be permanent copyright lobbyists, too.

Where people listen

Virtualmusic have attempted to track the various sizes of traffic to American-facing online music services in a heat map. This is what they came up with:



Click for bigger, and to see the methodology used.

Red services are shrinking year-on-year, green growing. And there's a few areas left out - YouTube, whose music videos create a box three times the size of everything put together; iTunes can't count because it's not a website; and MySpace because it's impossible to quantify which parts of the MySpace traffic is heading there for music.

Interesting things to note includes that Last FM has seen an uptick - despite having become a lot less flexible in how you can listen; AOL and eMusic have also grown audience in difficult years, and the figures here suggest that MOG is in the sort of decline that content points to.

Information Is Beautiful: Life on the minimum wage

Information Is Beautiful - the site that takes numbers and makes them pretty has pulled together a graphic which shows the activity a musician would need to do each month in order to make the US minimum wage.

It runs from selling own-produced CDs (143 at $9.99 a piece) to plays on Spotify (4,549,020 plays at $0.00043 a play).

Unusually for an Information Is Beautiful graphic, it actually obscures the facts rather than makes them clearer. For a start, there's a muddle between units - physical albums, single downloads and single streams is a bit like comparing the sales of cups of Starbucks coffee, packets of ground coffee and jars of Nescafe. An album is a bunch of songs, a single track is a single track, and to compare like with like perhaps the graphic should start from the premise 'how many songs must an artist shift' to make minimum wage.

There's also the problem of the way the data is portrayed - a massive circle for streaming, which makes it look like it takes a lot more effort than just selling a dozen CDs each month.

But that's a bit of a misrepresentation - it's hugely unlikely that any one person is going to buy your album more than once. If they love your song, they may stream it every day.

If someone in Arkansas likes one of your songs, they might tell a friend in Ohio about it. If you're using a business model based on selling self-produced CDs, you're going to struggle to be able to convert that into an instant sale. If your model has Spotify streaming, a link could be sent via email, and a stream play racked up there and then. (Assuming Spotify ever get round to launching in the US.) In effect, you're earning money as soon as someone says 'I think you'd like this artist' in a way that you don't get if you're only using physical sales.

There's also the marginal costs of production. It would be unfair to expect the graphic to show this clearly, but to sell 100 albums takes investment in producing 100 albums. Each extra album you want to sell costs a little bit. If you're incredibly lucky, you might get your stock levels spot on. If you're not, you'll either spend money on records you cant sell - and wind up like Alan McGee, hiding Legend! discs under the bed, eating into your profit margins - or you'll have demand you can't satisfy and lose money you could have earned. Streaming, once you're on the system, there's no marginal cost of production.

It's interesting in itself, but it only tells a portion of the story. It'd be nice to see a similar image with, say, the promotional costs needed to achieve these figures factored in.

[Thanks to Michael M, who suggested this would be interesting]

Spotify economics

Today's Guardian business pages have an interesting but limited piece on Spotify this morning; one which contains two statistics which are supposed to be impressive but, erm, actually aren't.

Number one:

But last week the blogosphere was alight with speculation about Spotify's model, fuelled by reports in its home market that it paid Lady Gaga – through the Swedish performing rights society STIM – a mere 1,150 kronor (£100) for a million streams of her song 'Poker Face'. Spotify stresses that payments to STIM only represent a fraction of the money received by rights holders and the figure is only for one country. But those reports have raised questions about how much money the music labels are making from Spotify and whether they have an ulterior motive for supporting the firm.

A million people listen to Lady GaGa, she makes a hundred pounds, and this is supposed to be a small amount?

As I've pointed out before, the last published figures showed that Radio One pays £18 a minute for music it plays - so, a three-minute Lady GaGa song on Chris Moyles show garners £54 for the equivalent of seven and a half million streams. Looks like Spotify is overpaying, doesn't it?

Number two:

There is embryonic evidence in Sweden that the rise of legal, free music services are helping switch people away from pirate sites. According to industry body IFPI Sweden, music sales are up 18% in the first nine months of this year.

Except the levels of file-sharing traffic in Sweden have returned to where they were over the course of the months since the Pirate Bay trial, so it seems unlikely that this boost has anything to do with people 'switching' from "pirate sites". Oh, and music sales are growing all over the place - British single sales are now over double where they were in 2002, for example, which is more down to the spread of usable digital services and the existence of singles people want to buy rather than any expensive file-sharing battles.

BPI claims 2% of BT customers are unlicensed file-sharers

The BPI has started moaning. They say they gave BT details of 100,000 customers they claim are file-sharing without permission, and BT haven't done anything with the information.

100,000? BT only has 4.8 million customers, which means the BPI is claiming 2% of their customers are filesharing to a point where something must be done. (You'll recall there's been suggestions that the music industry would only be interested in pursuing "heavy" uploaders rather than light users.)

BT have said that, much as they'd like to help, that many IP numbers are quite expensive to deal with:

Petter said he fears that the anti-piracy process could cost ISPs a staggering £365m a year – £165m a year more than the £200m the BPI says the industry will lose to online music piracy in 2009. The BT boss went on to label the BPI’s losses assessment as “melodramatic.”

Well, if the BPI are going to send so many demands, the figure of £365million isn't too extravagant - presumably Carphone Warehouse and Virgin customers would be roughly as evil, so we'd be expecting around a third of a million contacts to have to be made. That implies quite a few full-time jobs, and a lot of administration. The BPI isn't having a bit of it:

Geoff Taylor, chief executive of BPI, is now hitting back, claiming that Petter has exaggerated his figures too.[...]
“It’s shameful for a company like BT to know that a high percentage of the traffic it carries is illegal material but do nothing,” Taylor told The Mirror. “If you operate a commercial service and know it is being used to break the law, taking steps to ensure it is used legally is a cost of doing business.”

Is it? Really? Or is that rubbish? Is there any other business which expects someone else to fund the cost of its own security measures?

[Thanks to Michael M for the link]

Mercury Prize: Not a guarantee of big sellers

A handy chunk of data has been made available by the Guardian, in spreadsheet take-it-and-mash format, detailing not just the winners of the Mercury, but also how well they sold and the highest chart position.

So, just three Mercury albums have been number one - Suede by Suede, Different Class by Pulp and Whatever People Say... by the Arctic Monkeys.

Only three have sold more than a million - Franz Ferdinand, Pulp and the Monkeys.

Only one album failed to make the top 40, and that's also the lowest-selling album of them all: Talvin Singh's OK.

So, we now know that the average sale of a Mercury Prize Winner is a bit over 500,000 - 580882, to be precise.

And the average chart position is 10.

A top ten record and half a million sales? It sounds like the very model of a tolerably well-performing record. Not too flash, not setting the bar so high that the performance of the next record will disappoint, but good enough to guarantee the label will pony up for it.

Seven million filesharers? More or Less

The good people at More Or Less have been having a poke around in the figures for filesharing cited by Mandelson and chums to justify the clunking fist.

It turns out - after they dug and dug - that the figure is based on a report commissioned by the BPI. So, not entirely impartial, then.

Oh, and the survey size was quite small. So it turns out these 7 million filesharers are based on the confessions of just 136 people. And they were only 11.6% of the sample, so the research company just whacked the figures upwards:

That 11.6% of respondents who admitted to file sharing was adjusted upwards to 16.3% "to reflect the assumption that fewer people admit to file sharing than actually do it." The report's author told the BBC that the adjustment "wasn't just pulled out of thin air" but based on unspecified evidence.

How would you even be able to have a solid figure for making that assumption?

Oh, and also, the seven million figure is not only based on deciding arbitrarily to boost the numbers of filesharers, but by also adding an extra six million people to the numbers online - Jupiter, who did the research, estimated the size of online Britain as 40 million people, but at the time, the official estimate was just under 34 million.

If you adjust the figures to reflect the truth, then the seven million filesharers turn out be closer to 3.9million. Still, on the bright side, the Government can now claim to have helped reduce the figures sharing unlicensed files by almost half.

Bookmarks: Some stuff to read on the internet - Music industry figures

Thanks to Anthony T for pointing me in the direction of today's Ben Goldacre Bad Science column, which explores this week's claims that "researchers found more than a million people using a download site in ONE day and estimated that in a year they would use £120bn worth of material."

The whole thing is worth reading, but as a taste, Goldacre follows the trail to find out where the claims for "losses" come from:

What is the origin of this conservative figure? I hunted down the full Ciber documents, found the references section, and followed the web link, which led to a 2004 press release from a private legal firm called Rouse who specialise in intellectual property law. This press release was not about the £10bn figure. It was, in fact, a one-page document, which simply welcomed the government setting up an intellectual property theft strategy. In a short section headed "background", among five other points, it says: "Rights owners have estimated that last year alone counterfeiting and piracy cost the UK economy £10bn and 4,000 jobs." An industry estimate, as an aside, in a press release. Genius.

Pete Waterman: How did he reach 154 million?

Pete Waterman has, so far, been the biggest moaner taking the PRS line on YouTube royalties. You'll recall his complaint:

"If 154 million plays means £11, I get more from Radio Stoke playing Never Gonna Give You Up than I do from YouTube."

Now, we've already pointed out that he was including views from all over the world - not just the UK - and that he only qualifies for a writer's payment, and that has to be split three ways.

And we've always had our doubts about 154 million plays in the first place.

Handily, the good people at Visible Measures have launched their list of the most-watched videos of all time. These figures are interesting because Visible Measures use True Reach, which counts spoofs, forwards and derivative works as well. They reckon 18 videos have broken the 100 million mark.

Rick Astley doesn't appear on the list.

Now, they're counting in one way, and there's room for margins of error and so on. But there's a massive gulf between "154 million" and "less than 100 million", don't you think?

YouTube refresh their approach to viewcounts

Not entirely unconnected with the current debate over how much money bands should get for their videos being shown on YouTube comes an announcement from the tubers that they're getting tougher with bands and others who get their viewcounts artificially inflated:

Video view counts reflect the YouTube community's interests and the grassroots popularity of videos. We periodically make changes that allow us to display consistent view counts and accurately reflect a "real" view based on video consumption, video streaming and spam filtering. Unfortunately, a few people still try to artificially manipulate their video's view counts. Some people game third-party view counts as well. That can make things unfair for everyone.

Recently, we found spamming issues associated with the view counts on a small number of videos. The inflated view count number on these videos will be frozen until actual views catch up to the published, artificial, view count. Also, a few people have commented that their view counts are updating more slowly. Occasionally the speed with which views update changes -- sometimes it’s faster and sometimes it’s slower. But we are always working to make sure that the final view count numbers are an accurate reflection of the community's interest.

In other words: hitting F5 over and over again will do you no good.

Of course, there remains a slight question over how YouTube will determine if its system is being gamed or not; you'd hope that - given these figures will be at the heart of any recompense to artists - they'll share the data upon which these decisions are taken. Otherwise it could all look a little high-handed.

IFPI: Adding up the downloads

That IFPI report? The NME runs a chunk of the findings:

The International Federation of the Phonographic Industry (IFPI) Digital Music Report 2009 estimates that 40 billion files were unlawfully shared online in 2008 alone.

The report states that the IFPI removed three billion links to such files in 2008, up from 500,000 in 2007.

Actually, the report states three million, which is a little more credible.

And do the IFPI really mean 40 billion files are being "unlawfully shared", or are they adding up the number of times a smaller number of files have been accessed? You have to actually go to the full report to find out. The advantage, of course, is that you do get unfunny cartoons as well.
So, what do they actually say about this forty billion files?

Estimates on the impact of internet piracy vary but are consistently huge in scale. IFPI, collating separate studies in 16 countries over a four-year period,
estimated unauthorised file-sharing at over 40 billion files in 2008. This means that globally around 95 per cent of music tracks are downloaded without payment to the artist or the music company that produced them.

Hang about... this isn't just any old made-up figure, then, but one that has been hammered together from just sixteen countries, a bunch of studies which will use a range of different methodologies and from over a four year period. Which is more or less forever in internet terms. So even if it wasn't a made-up figure, it would be a totally pointless. The IFPI doesn't provide any further indication of how it made-up the number, nor even the studies on which it built its figure. It doesn't even define its terms, so we don't know if there are forty billion supposed downloads of unlicensed files, or if there are forty billion files available for download and possibly being sucked off the internet hundreds of times.

The fallen, phantom 750,000

At a time like this, when the doughty custodians of public money turn out to have not grasped that high rates of interest go hand-in-hand with increased risk, and ITN has sent someone up into the loft to see if they still have that UK Job Losses Map they used to track Thatcher cutting a swathe through British manufacturing, any threat to the economy has to be taken seriously.

Like, for example, the oft-quoted figure of 750,000 jobs that are lost in the US to piracy. Any threat to a quarter of a million incomes has to be taken seriously, right?

Only it turns out - after ARs Technica do a bit of digging - that this figure has no basis whatsoever; like a Wikipedia entry referencing an article which was based on the Wikipedia entry in the first place, it is supported merely by reference unto itself, perhaps sparked by a finger-in-the-air estimate made by Commerce Secretary Malcolm Baldridge. That would be Reagan's Commerce Secretary. He was talking in 1986. Oh, and 750,000 was a top-end, on a range he proposed that stretched from 130,000. Take the lower estimate, and you've successfully re-employed 620,000 people. Job creation really is that easy.

Likewise, the USD200-250 billion dollars that is sucked from the economy? That appears to have come from a 1993 edition of Forbes, and referred to all counterfeit products. Oh, and that was worldwide. And when Napster wasn't really much of a threat. It's a bit like Sarah Palin using back numbers of Punch to try and create a foreign policy position.

It's possible, of course, that the damage to the economy might be harsher than these white-bearded old numbers suggest - the entertainment industry's reluctance to think about the numbers could, humorously, be actually making its case less compelling. We don't know. More importantly, though, neither do they - and they're calling for changes in the law on the basis of made-up numbers.

[Thanks to Michael M]

Gordon in the morning: Journalism

We're sure that - once his hands were both free - Gordon Smart punched the air as he took delivery of topless pictures of Lily Allen. He even created a new euphemism for the occasion:

Lily got something off her chest and peeled off her shirt to reveal a couple of chart-toppers.

Meanwhile, Gordon has been carrying out a poll. An online poll. He publishes the results without any warning that the respondents were self-selecting and, therefore, not scientifially accurate. But it might not matter:

2 out of 3 men fancy Posh Vic

He has done a pie chart, though.

Radiohead play coy on size of pot of gold at end of In Rainbows

Comscore's finger-in-the-air claims that about a third of people paid for In Rainbows seem to have become solidified as fact now, with the only debate whether getting more than one in three punters to pay cash for something they could get for free is a good result, or having nearly two-thirds of customers take your goods without so much as a spit in the spittoon is a terrible loss.

The figure is almost now gospel, despite Radiohead's attempts to deny it:

As the album could only be downloaded from the band’s website, it is impossible for outside organisations to have accurate figures on sales. The figures quoted by the company comScore Inc are wholly inaccurate and in no way reflect definitive market intelligence or, indeed, the true success of the project.

Which is in itself definitive, but surprisingly lacking on substance. Of course, Radiohead could blow Comscore out the water by releasing figures themselves, but they choose not to.

Which, of course, they're perfectly at liberty to do - although it hardly fits with the band's attempts to distance themselves from the pie charts and sliderules of the music industry.

Indeed, while Radiohead has been scrapping with Comscore, one of the big, old-style companies has turned up to hold the coats:

Warner Bros Records senior technology director Ethan Kaplan quickly savaged comScore’s methodology, which included only a “few hundred” in the study.

Oddly, Kaplan has never complained that the Grammys crown "best" artists and records based on a tiny sample, but never mind.

Comscore haven't taken this slight lying down. They've powered up their blog engine to launch a defence of their counting and stuff, and the idea of sampling in general. Who knew that Radiohead's album experiment would lead to statisticians rowing?