A client (a major orchestra) recently asked me about a post by Norman Lebrecht about classical charts. The orchestra in question was concerned about what appeared to be a precipitous decline in sales, and wondered if it was sensible to embark on a recording project under these circumstances. I had to spend an evening undoing the damage that Norman’s idiotic bullshit had done.

Unlike Norman, I run a successful record company, so I generally have better things to do with my time than correct his ill-informed musings (if you want to see his ethical and epistemological failings catalogued in relentless detail, I suggest you follow @fakenlebrecht). Still, I’ve had to write this now, and I might as well share it.

Each week, Nielsen in the US and the Official Charts Company in the UK circulate charts showing the top selling new titles in each genre. They don’t publish the sales figures because it’s commercially sensitive information. “Giving it to inept egobloggers” is not on the list of the things you’re allowed to do with it. Despite this, the spreadsheets are circulated fairly widely, and Norman gets hold of them from time to time. Reading his commentary, you could be forgiven for thinking that a continuing trend of declining classical music sales had reached a new low.

Indeed, it’s hard to read it any other way. That’s because it is based on the most idiotic interpretation of the data I can imagine.

To come up with this interpretation from the actual figures, you’d have to ignore a significant amount of publicly available evidence to the contrary, you’d have to have almost no access to real record sales data, you’d have to be unfamiliar with the way charts work, make barely any effort to find out how they work and, crucially, to have no regular access to them.

Here’s how it works: charts include only a subset of recordings, count only a subset of sales, the sales of the #1 title vary hugely both seasonally and from week to week and are an extremely poor indicator of the health of the general business.

Every week, the official charts company emails me the UK’s specialist classical chart, which lists sales of albums less than 12 months old comprised of more than 60% classical repertoire (by length) as reported by participating retailers.*

Here’s a graph showing the sales of the #1 and #20 album on this chart, each week for a year. You could combine this with the charts themselves to get sales figures for all the best-selling albums of the last year, so the OCC asked me not to label the Y axis. That’s reasonable, since we don’t need it to understand the point. I can tell you that several of the #1s on this chart are records I’ve worked on, and one of those sold almost 1,000 units in a week. The US charts look similar, but the numbers are bigger.

There are easily more than 100k classical albums available to customers in the US and UK**. If you wanted to know how they, collectively, were selling, you could look at figures published by Nielsen and the BPI, which consistently show that while the entire record business has halved in size over the last fifteen years, classical sales are basically stable and not currently in the midst of a precipitous collapse.

If you insisted on using the sales of a single album in a single week as a proxy for the success of the business, the smart money would be on picking any album except #1. It’s literally the worst way of doing this.****

==

* That is to say, if a record is more than a year old, was only sold by a retailer that didn’t report sales to the OCC, or wasn’t deemed classical enough, it could sell a million units and not appear on this chart. I don’t know of anybody shifting a million units a week through museum gift shops and CD signings, but I’ve also arranged to sign up a retailer to the OCC’s reporting panel because we once missed out on having the #1 classical album on the busiest week of the year because we sold hundreds of units in the wrong shop.

** Citation needed? Ok. On their website, Arkivmusic.com claims to have more than 120,000 recordings available in the US. I called Chris O’Reilly*** at Presto Classical in the UK. He told me he thinks they have almost everything available, with 78,000 core classical titles on CD and DVD, and another 40,000 digital-only audio products. My experience tends to suggest that Chris is basically right about core classical repertoire in the UK. If you count compilations, crossover albums, and little labels without distributors, there are probably a few more than this. The total is perhaps as many as 300,000 individual titles on sale at any one time, from thousands of labels around the world, and a subset of maybe 150,000 available in major markets like the UK and US.

*** You can just call people who know things and ask them stuff, and then check that against another source to see if it’s likely to be true. This seemed infinitely simpler (and plainly more reliable) than whatever journalistic technique resulted in the following story, every word of which is untrue:

**** If you care about the truth, that is. If you’re totally cool with wilfully deceiving your readers and undermining the factual basis upon which they make business decisions, you will find constructing false narratives around isolated statistics to be the gift that keeps on giving. Which might be why it keeps happening.