Moses Supposes: RIAA and SoundScan clash on sales stats

Moses Avalon is one of the nation’s leading music-business consultants and artists’-rights advocates and is the author of a top-selling music business reference, Confessions of a Record Producer. More of his articles can be found at

The following includes slides and facts presented at my NAMM 2011 panel presented by H.O.T. Zone as well as reprinted material from a 2004 article.

In April of 2004, I was working on a piece for Keyboard magazine about bar codes. I was trying to determine if the “free” bar code that many CD replicators provide is a real added value to the indie artist, or just a bogus premium.  But what evolved was a much bigger and more important story.  Talking to SoundScan revealed a massive untruth disseminated by the RIAA about the reality of declining record sales.

SoundScan (owned by Nielsen) uses the bar codes on CDs to register sales at record stores.  The correlated data contributes to the Billboard chart listings, as well as much of the market research used by record companies.  Through my investigation, I learned things that would contradict reported statements by the RIAA — mainly that US labels had a significant reduction in sales from 2000-2004.

My panel at the 2011 NAMM show this year is all about getting to the bottom of the fallacy that the music business is in disarray, and this means understanding why SoundScan can say that sales are stable yet the press — the RIAA — can say that sales are plummeting.

So to facilitate that, I’ve reprinted the key portions of that April 2004 article in this piece.

RIAA’s “Losses”

Cary Sherman, president of the RIAA, responded personally to me in April of 2004, and, in the process, exposed intriguing insight into the way the RIAA calculates “losses.”

In early 2004, Sherman claimed in a speech to Financial Times Media at a broadcasting conference in London that the industry was experiencing a “7% decrease in revenue since last year [2003].” He named piracy / file sharing as the main reason.

But when speaking to a representative from SoundScan, I learned the following:

– For the first quarter of 2003, SoundScan registered 147,000,000 records sold.

– For the first quarter of 2004, SoundScan reported 160,000,000 records sold.

That’s 13,000,000 more units, almost a 10% increase from 2003, at the time Sherman made his assertion that sales were in decline.

So I asked the SoundScan rep, who would only speak to me if I didn’t use his name, “Would you disagree with what the RIAA is implying?”

“I would never disagree with the RIAA,” he said, smirking.

Forget the confusing percentages; here’s an oversimplified example: I shipped 1,000 units last year and sold 700 of them. This year I sold 770 units but shipped only 930 units. I shipped 10% less units this year. And this is what the RIAA wants the public to accept as “a loss.”

But he did explain the rationale that would allow both of these seemingly inconsistent realities to exist in the same universe: “The RIAA reports a sale as a unit shipped to record stores, whereas SoundScan reports units sold [to the consumer] at the point of purchase. So you’re talking about apples and oranges.”

Really!? I fact-checked this with Cary Sherman, who confirmed, “He is correct,” and added, regarding RIAA and SoundScan data, that “the two sets of numbers tend to be similar, but because of timing differences, they’re usually a little different at any point in time.”

Similar?!  How is a 10% increase for first quarter of 2004 similar to a prediction of a 7% decrease for the entire year of 2004?

The Secret: “Shipments” = “Sales”

Now armed with the secret-decoder formula, I went back and read the RIAA and International Federation of the Phonographic Industry (IFPI) websites more adroitly.  Sure enough, every time the RIAA complained of large drops in “unit sales,” it included international sales, not strictly domestic.  Every time it spoke to domestic “losses,” it was referring only to “units shipped in the US” to record stores. It seemed obvious that if the RIAA confined its revenue statistics to the US market alone, they may not be able to publish any losses in revenue at all for 2000-2004.

There is only one logical integration of all these statistics with the SoundScan data: even though actual point-of-purchase sales were up in the first quarter of 2004 by about 9% in the US (and the industry sold over 13,000,000 more units in Q1 2004 than in Q1 2003), the industry is still claiming a loss of 7% because RIAA members shipped 7% fewer records than in 2003. (Sales would ultimately level off for 2004 at a 12% decrease.)

Forget the confusing percentages; here’s an oversimplified example:  I shipped 1,000 units last year and sold 700 of them. This year I sold 770 units but shipped only 930 units. I shipped 10% less units this year. And this is what the RIAA wants the public to accept as “a loss.”

In other words, had major labels been more conservative and less pathologically optimistic about sales growing each year, they would have shipped about the same amount of units and then the RIAA would have nothing bad to report.  Sales of albums (remember that in this comparison we’re only talking about albums) leveled off slightly during between 2000-2004, mostly due to the fact that cassettes were all but discontinued, and digital downloads of singles became the fad.  (See the chart below.)

I’ll go a step further. Fewer shipments and more sell-through should mean higher profit margins and faster turnaround; and shouldn’t that be good for both the retail and wholesale side of the industry?  “Sure,” admitted Sherman.

Album Sales 1985-2010
Actual album sales only dipped a bit between 2000-2004 and were on the rise during the height of the so-called Napster debacle. From 2004 to today, we see decline, but this chart does not account for singles sales or increased revenue from licensing. Such figures would show that overall revenue is up about 20%.
Labels' Costs 1985-2010
Labels are spending almost 30% less to sell the same amount of recordings since 1985. Where is all the profit going? See the next slide.

So Are There Real Losses?

Maybe, but “we, the people” will never be able to figure them out due to this confusion, deliberate or not. Regardless, it’s certainly been a great excuse for majors to clean house of overpaid executives. As for overpaid CEOs…well, see for yourself.

Salary Structure for Major Labels 1985-2010
It's easy to say profits are down when you keep paying yourself more and your employees the same for 20+ years.

But as for a US major label’s bottom line, the effect could never rise to the RIAA / IFPI claim that file sharing is the “major factor” of revenue loss for labels, and certainly not for artists — at least not before the advent of LimeWire in 2004.

Nope. My analysis suggests that the number-one reason for the loss of jobs in the industry is self-perpetuating major-label PR, and that the number-one cause of loss of unit-sales revenue for artists is still record-label accounting practices.

If sales are rising or leveling, but profit is falling, there is only possible explanation: somebody in control is being paid a lot more now than they were in the good old days.  Who could that be?  The senior VPs?  No, their salary has stayed about same over the past 15 years at about $300-$350K/yr.   Entry level interns?  No, they have kept pace with inflation rising from about $18K a year to $35k, just past the minimum-wage level.  How about CEOs?  Yep.  The average CEO salary has bolted from $850K in the mid- to late-1980s to a whopping $3 million a year in 2010.

Looks to me like we found out what is really causing declining profits, and it ain’t the cost of polystyrene.

Leave a Comment