Music Discovery and Purchasing Survey Results Part 1

We conducted a survey of how a small group of 187 self-selected respondents discovered and purchased music.  I’m publishing these results because they were interesting and may point to some ideas about how more rigorous surveys should be undertaken.  There will be one more post with the rest of the slides.

What Kind of Stream Am I?

One interesting trend in the survey is that consumption charts may be measuring the wrong thing or measuring it the wrong way.  Spotify has suggested that Spotify streams should be valued at 100 streams to the value of a single track download.  Billboard has counted 1500 streams to a single album download at a $7.50 wholesale price.

[D]igital interactive streaming — not including passive streams from services like Pandora, iTunes Radio and Sirius — appears to be making up the [year over year decline in digital revenue], on a revenue basis at least. According to SoundScan, interactive music and video streaming totaled 34.28 billion streams in the first quarter of the year, versus 25.44 billion streams in 2013’s corresponding period. Not only have streams grown by nearly 9 billion, but per-stream payout has improved this year versus last year, according to label sales executives.

Last year, interactive streams paid an average of $0.00375 per stream, meaning 2,000 streams equaled the average $7.50 wholesale — the average price when you consider the $9.99 list price for most albums and $11.99 for some superstar albums — for a digital album download. This year the industry average is more like $0.005, which means that 1,500 streams equal the wholesale cost of an album.

Figuring 2,000 streams per stream equivalent album last year, and 1,500 streams per stream equivalent album this year, that means that SEA totaled 22.85 million in the first quarter of this year, while last year totaled 12.72 million, a difference of 10.1 million album units. So, while digital albums and track equivalent albums were down 9.06 million units, streaming revenue growth is outpacing digital sales decline.

First, in discussing the value of streams, it’s always important to distinguish which kind of streams you are counting.  Ad supported streams are of vastly lower economic value than subscription streams, so it’s probably not consistent to count “streams” without being very clear about which stream you mean.  The variability in accounting for streaming revenue is compounded by having (for the most part) no fixed rate on streaming plays due to revenue share deals which will vary from accounting period to accounting period in both the nominator and denominator.

Slide 2 suggests that there may be questionable value in streaming services like Pandora and YouTube as music discovery platforms which may make sense since both services involve searching for something you already know about or play me something like something I already know.

Again looking at trends, there is a large majority of respondents who spend $10 or less per month on streaming services and a plurality of respondents who spend $10 or less per month on streaming and permanent ownership formats combined.

Not to be overlooked, 12% of respondents said the did not buy music at all.

We will be doing a more detailed live music survey, but note that a significant plurality of respondents spent less than $25 per month on live music.

The “Zero Effect”:  Do Consumption Charts Penalize Artists Windowing Streaming Services and Most Compilation Records?

Although we can’t tell how consumption charts are weighted (as far as I know that information hasn’t been released publicly), it’s pretty clear that if you are not on streaming services–meaning you have zero streams–you will be penalized in the chart.

Based on data for the week ending November 2, 2014, Taylor Swift’s first week sales were so strong it wouldn’t matter that her streams were somewhat lower for the consumption chart.  At #1, she outsold the #2 NOW 52 title by 10:1, and NOW 52 outsold the #3 Sam Hunt album by 5:4–but Sam Hunt had 4 million streams that punched up the chart position.

The “zero effect” is much greater further down the chart, however.  In the same week, “Frozen: The Songs”, a compilation record, got credit for zero streams and 10,723 albums sales for a chart position of 49.  Blake Shelton sold 8,735 albums but got credit for 930,928 audio streams and a chart position of 44.  The same week Iggy Azalea sold 4,947 albums but got credit for 5,060,617 streams for a chart position of 25.  In other words, Frozen will never have any streams because compilation records typically do not get streaming rights, and got a much lower chart position in spite of selling over twice as many albums.

If you compared based on album sales along, Guardians of the Galaxy zero effect soundtrack would have entered the chart at #25, not #40, Sam Smith would have been #15 instead of #6, Bob Segar would have been #23 instead of #34.  Another zero effect soundtrack is Now Disney 3 that would have been #40 instead of #59, and U2’s Songs of Innocence would have been #64 instead of #94.

Seasonal records such as Christmas albums are also penalized.  The Nov 2 chart showed that based on album sales alone, Home Free’s Full of Cheer would have entered the chart that week at #66 instead of #104.  While the title had 26 streams, that was a sufficient penalty to cost the record 38 chart positions.

Conclusions?  Charts are relative beasts to begin with, and the consumption chart won’t keep a phenom like Taylor Swift from dominating the top position.  Measuring streams probably isn’t enough to affect the top 10.  But for records that are compilations, soundtracks, seasonal or other specialty titles that either aren’t allowed a streaming audience based on contract, are windowed, or haven’t found that audience yet for another reason, the consumption chart penalizes high sellers that are not present on streaming services.

If chart position matters to your record, then this should be of concern to you as the zero effect creates an incentive to stream.  Some would say that the more streaming, the lower the sales.  Without getting into cause and effect on that issue, it certainly can be said that the lower the streams, the lower the chart position even if sales of a given artist are higher than another given artist.

From a profitability perspective, artists whose records sell but don’t stream may well be thankful.  If that trend continues, then it would also stand to reason to question the benefit of chart position as a selling tool.  But then we hear about services like YouTube routinely deleting billions of fake plays in its video playlists during December.  If this same phenomenon is repeated in streaming services used to measure chart position….not to imply that anyone in the music business would ever try to rig the charts.  Perish the thought.

So what is it all about?  Is there a “zero effect” or is there zero affect?  Sales or streams?

1.  Age

Age 1

2.  Music Discovery from specified sources

Music Discovery3.  Monthly spend on streaming services

Monthly Spend on Streaming Services

Comment:  Based on this slide, it seems unlikely that respondents subscribed to more than one streaming service.

4. Monthly spend on streaming plus CDs, downloads or vinyl

Monthly Spend on Streaming plus CDs downloads

A near plurality spend less than $10 a month on combined sources, similar to the purchasing on streaming only. The majority spend more that $20 but less than $50.

5.  Music buying location

buying location

bit torrent prelim

The 3rd bar on the Bit Torrent slide is “I don’t know a retailer called Bit Torrent”.

6.  Format purchased


Downloads nearly equal CDs and vinyl.

7.  Live show spending

Live Shows 1