By George Ziemann - 10/22/10
A new school year is underway and it's time to give all the college researchers something new to chew on.
I've been studying the recording industry statistics for 8 years, a move which was prompted by the RIAA interfering with my attempts to sell my own music on eBay in 2002. Over those 8 years, a brief instant in the history of recorded music, the entire game has changed and it will never be the same again.
There are a few numbers to show, some data to analyze, but the real story cannot be found in any industry reports. In fact, the more I look at the labels' year-end reports, what the RIAA and Nielsen put out, it becomes more obvious that it's all pretty worthless. Not one of them holds the real story. Not even a fraction of it.
You can memorize every fact and figure from Nielsen's data and still not know a damned thing about what is happening in the music business.
But the truth is out there. And once you start putting the pieces together, the big picture is hard to ignore.
My original vision of this article was divided into several sections, each with a specific focus. But that's because I was trying to be all academic about it. All that does is needlessly confuse the issue. It's really very simple, so I'm going to keep it that way.
U.S. Releases and
New Releases -- U.S. (1992 to 2009)
I'll leave you to ponder over that for a minute. Now let's look at Canada, some Statscan data provided by Kent Clark, a Loyal Reader from the Great White North.
The Canadian data is rather odd. It doesn't follow the cause and effect model either.
Correction from previously stated assumption -- "Other" is data from non-Canadian releases.
The Canadians slashed their new releases between 2003 and 2007, and sales actually went up. Maybe the Canadian branches of the major labels actually got more efficient at connecting with the audience. Maybe Nettwerk Records has a lot to do with it, too.
Here's the big difference between the U.S. and Canada:
When I started writing about the RIAA in 2002, nobody had a clue who they were. I had to always explain it. Now, pretty much every college student in the country knows all about them. I don't have to explain it. The RIAA is the group of assholes that sued 40,000 people for listening to music. And every year since they really got going, sales have dropped by 15 to 20 percent. What a big surprise.
Canadians haven't been sued. Yet. Nettwerk Records, a Canada-based label, actually came out against the idea of suing people, which is why I think that may have effected sales in a positive manner.
There's still a cause and effect going on, but it's based on knowledge, not numbers.
The thing is, all of this is such a miniscule part of the actual music industry, especially in the United States. Nielsen is considered to be the most reliable infor mation. It's what people turn to when they want to find out what's happening as far as data goes.
But Nielsen doesn't have half the story.The Unreported Sales
If you add up all the available new release figure from 1992 to 2009, you're going to come up with a total in the neighborhood of 735,000 new releases.
Go take a look at CDBaby's "About" page. According to CDBaby, they are "the largest online distributor of independent music." And they'll give you their numbers.
278,510 albums being sold on CD Baby
5,339,025 CDs sold online to customers
$107,769,092 paid directly to the artists
The stat there that I'm keying on is 278,510 albums being sold. That's a full third of what Nielsen reports to have been released since 1992.
In 2005, Tunecore arrived (Note). They're a little less public with their numbers, but I know that last year, they claimed 90,000 new releases, just about the same as Nielsen reported for the entire industry. Unless CDBaby went on a hiatus for the year, obviously there's a huge pile missing (CDBaby is going to provide their 2009 numbers, which will be added here when then arrive). Let's add Catapult to that. They released more than 10,000 new titles in 2009. There's also Ditto Music, IODA, The Orchard, MusicBizAcademy, RouteNote, FUGA, and that's just the first page of Google results for "digital music distribution." Who's missing from this picture? How about ReverbNation? If you use FaceBook at all, you know who they are. They wouldn't give me any numbers ("...we just don't release numbers about our premium services as a matter of policy"), but they are certainly one of the major players in independent digital distribution.
And what about the majors and what they distribute?
Makes Nielsen's numbers obviously useless to determine what's really happening. So what is happening?
I talked to Derek Sivers, who sold CDBaby off to DiskMakers around 2007-2008. He said that when he left, they had been taking orders for around 200 new album releases every day. That's more than 70,000 new releases in a year, which is just slightly under what Nielsen shows for the entire industry at that point in time. I don't know how many actually were added to the stores. Derek couldn't tell me that, but did say that the majority of them were orders for 100 to 1000 copies to be duplicated.
What happened after these copies hit the streets and landed in the laps of the musicians that ordered them?
And what about all of us with the time, patience and means to simply burn our own CDs? I'm using Tunecore to place our music on iTunes (Hayden's Wall and Hurricane Alley -- go buy a damn copy so I can pay the electric bill). But the truth is that we sold about 50 copies off the edge of the stage for every one that we sold online.
Nielsen never counted any of them. Or any of the CDs that the other thousands upon thousands of independent acts sold off the stage. Hell, most of them probably didn't even report it on their taxes (Dear IRS -- I did).
The Big Question
While ReverbNation didn't provide me with their data, Jed Carlson (co-founder and CEO) did have a very pertinent question to ask. He wasn't the first to ask it, but I like the way he phrased it:
As a reader, I'd like to know:
1. How many album releases (and how many total songs released via album format)
2. How many single releases
3. How many video releases from each distributor AND from soundscan.
Just seems to me like there is a lot of murkiness in the way people are choosing to define words like 'release'. Need to compare apples to apples.
"Murkiness" doesn't even begin to describe the music industry's approach to their data. When I first started discussing new releases in 2002, the RIAA immediately removed all new release data from their website and it took BusinessWeek to get an answer from them (the RIAA has never answered a single inquiry from me), which conflicted with what they had previously announced in press releases. They have since moved all their data behind a paywall. And Nielsen charges a fortune for their numbers, which are obviously inaccurate.
A representative from on of the other distribution services made the following observation:
I think they [SoundScan] are great at tracking labeled artists sales from brick and mortar and even their digital tracks from the major music stores - but when it comes to independent artists (Which make up over 50% of the purchased music market) they are lacking any real credibility.
Personally, I don't think the industry wants anyone to be able to compare apples to apples. They prefer to throw everything in a blender so that the individual components which make up the industry statistics are impossible to distinguish. It's so much easier to hide the truth when it isn't really there to begin with.
"What you'll see in the near future is a million artists and about 500,000 labels, and many artists becoming their own labels. The whole album paradigm will be redefined. It will be a singles marketplace and it will be shared by a million hands."
Chuck D. -- July, 2000
"A paradigm is like a theory, but a little different. A theory is an idea that sets out to explain how something works, then it's tested, proved or disproved, supported or challenged by experiment and reflection. A paradigm is a set of implicit assumptions that are not meant to be tested; in fact, they are essentially unconscious."
What the Bleep Do We Know, via FaceBook Sept. 2010
Since 2003, Apple has sold more singles than the music industry did over it's entire history prior to the iTunes Music Store. 10 billion songs, last I heard. There's still a problem with quoting this and keeping it in context, though. Many (perhaps the majority) of those song sales were as part of complete albums. Still, the industry had almost completely killed the single before Apple got involved with music sales, so even if it doesn't directly apply, it is significant in terms of the landscape shifting.
Services like Tunecore have been singled out for criticism by people like Tommy Silverman, who will tell you that 80 percent of the artists using Tunecore to get into Apple are crap because they sell 100 copies or less in a year.
Those are the people who are using TuneCore and iTunes to clutter the music environment with crap, so that the artists who really are pretty good have more trouble breaking through than they ever did before."
Silverman missed two very important points.
The first is that 80% crap is still a 20% success rate. The RIAA only claims 5% success. So that's 95% crap. And Silverman's record is no better.
Tunecore lets anybody through. The labels pick and choose. Tunecore has a higher success ratio. Draw your own conclusions on that one.
The second point is that as little as 6 years ago, none of those artists selling less than 100 copies were in retail at all. We sold zero copies online. So even one sale is better.
The thing I'm focusing on from Silverman's statement is this: "...the artists who really are pretty good have more trouble breaking through than they ever did before."
That seems a little ridiculous as a general statement. Just having an album (or 100,000 albums), on iTunes doesn't interfere with anyone's sales. They don't bury other releases (like the majors used to do with cover songs, where a minor hit would come along and all the big stars would record their version which would literally "cover" the original release at the retail level).
iTunes works kind of like peer-to-peer, in the sense that you generally have to know what you're looking for in order to find it. If you type in "Mariah Carey," that's what you're going to get. Nothing gets in the way.
Maybe the major artists have more trouble breaking through than ever before, but (in my opinion) that's because the RIAA taught the college students to hate them.
And let's talk reality. It hasn't been "the artists who are really pretty good" that have been at the top of the heap for decades. Not since disco hit. Now it's the ones who are willing to sell out for quick cash, follow the rules, fit themselves into the mold of the flavor of the month and whore themselves out.
Being talented has almost nothing to do with it. It's all about your ability to do business, not your skill at playing music. That's why we have autotune.
Back to the point, which is that today's music business simply cannot be seen in the industry reports. You can pile up all the stats and data and the real music business just isn't in it. It's out on the streets.
Times are tough. People have less money than ever to toss around on music. Nothing against Tom Petty (he's just my favorite star to pull out as an example), but if I get $10 from you for a CD sold off the end of the stage on Friday night, you might have to pass up on Tom's new album when you spot it on Saturday.
And there are hundreds of thousands of other acts doing the same thing, almost every night of the week (your mileage may vary depending on the size of the town). The artists "who are really pretty good" are selling records.
But they're not on the charts. The charts don't tell you anything.
In the last 10 years or so, from mp3s, CD-Rs and the internet, then the ability to get into the world's largest retailer while completely bypassing the traditional label system, the entire paradigm of the music industry has changed. No one sees it all. No one knows how to measure it. Everything is upside-down and off the charts.
It's a new day, a new beginning.
The beginning is a very delicate time. -- Frank Herbert, Dune
Notes and Follow-Up
From Statscan's Sound Recording and Music Publishing data tables for 2006 (via Kent Clark):
"It should be noted that data from this historical time series should not be compared with data from this new survey due to significant differences in coverage and methodology.
"The new survey covers a somewhat different set of businesses than in previous years so that data generally cannot be expected to be comparable. The list of names and addresses of businesses is now drawn from a central Statistics Canada data base. Also, a much more rigorous delineation of those companies that are considered part of the culture sector has been applied through the implementation of the North American Industry Classification System (NAICS). This industry-based classification is a departure from the activity-based classification that was used previously. In addition to these changes in coverage, commencing with 2005, the data are based on a sample of businesses."
So, the 2007 numbers are actually a different measurement than previous years. However, they did modify historic data to create 2005 numbers that could be trended.
"So I read it, and it sounds like you've given up on your old theory. I don't think you should. It's still valid, and its really obvious that the industry has continued to produce fewer albums."
It's not so much that I've given up on my old theory, it's just that the available data is in contradiction with the basic cause-and-effect that used to seem so obvious.
"This recent SoundScan number sounds like they're counting new data sets that weren't previously being counted."
In a way, that's entirely correct. When we had record stores, Nielsen used to survey some of them and estimate the rest. A huge number of independent releases that would have previously been ignored now go through iTunes, Amazon, etc., and they finally show up in the data.
"Also, the graph mixed with the soundscan stat is misleading, though I see what you're trying to do. The soundscan and the graph data are measuring totally different things, and placing the two pieces of data near each other implies a correlation."
My original theory drew a correlation between sales and releases. Yes, they are two completely different measurements. But there used to be an apparent correlation between them.
Note that some of SoundScan's data is intentionally misleading. For four or five years now, every time their stats come out at the end of the year, they've been touting an increase in unit sales (or "purchasing decisions"). The problem with that methodology is that selling 300 singles for a buck each ($300) becomes a 50% increase over selling 200 albums at $10 each ($2000).
I've seen this trip up many people who should know better.
"I was looking at your 2010 update on music sales and I found this chart at The Happy Hospitalist, and i'll be hot diggity darned if the peaks and valleys of the velocity of money in the US economy don't line up exactly with the units shipped chart on your page. OK, OK, it's not perfectly lined up, but people just didn't pick back up buying albums when the market rebounded a tiny bit. Also, as you point out, the RIAA was suing their customer base. Very interesting." Nathan Curry
1) The original version of this article stated that Tunecore started up in
1996. The correct date is 2005. Actually, when I wrote the article, I was under
the impression that it opened in 2006. It still would have been wrong, just
not as eggregiously wrong. I have no earthly idea why I typed in 1996.
©2002 George Ziemann
All material on azoz.com is protected by copyright law and by international treaties, but it all seems rather pointless.
You may reprint any article on this site in whole, in part, in effigy or in ridicule.
I really don't care.