The Brooding Savage

Readers of this blog know that I have reservations about the feasibility of ad-supported music strategies, like Imeem.  (For a look back at some of my earlier analysis, click here or here.)  But I have to defend the revenue model for businesses such as Imeem --  if they can manage to build an audience, they should be able to make a go of it.

Up until now, I have only weighed in on possible ad-supported models for actually distributing and downloading music files.  But a new opinion  penned by Silicon Alley Insider raises the question of ad-supported music streaming, too.  And, like I’ve also chimed, it’s largely the fault of music labels and the RIAA.

Now, my basic problem with streaming music services is that they’re basically the same as radio – which has always been free.  XM and Sirius tried pay radio (and I admit their product is far superior than FM) but we all know how poorly these businesses performed. 

As much as you’d like to think that streaming services like Imeem are better than radio – it just isn’t true.  A really good radio broadcaster actually has better service than anything users can program themselves.  As evidence, I site a recent Pew study that says Radio listening is how 83% of Americans discover new music.  In other words, the public relies on Radio for it’s ability to expose them to songs they don’t yet know – and this service goes right out the window if the service is solely an on-demand product.   

No. I’m afraid the price point has already been set in this streaming market.  And the price is nil.

(I have also argued that music is something that most of us buy with an intent to form a collection. I still have my old Todd Rundgren albums, even though I don’t even have a turntable – music after all is a reflection of not only who I am, but also who I was. The idea that I would pay money for music, and then not actually “own” it, is plain silly.)

But SAI’s issue is not with the product concept.  According to their report, the financials of ad-supported streaming don’t add up.  Here’s the math:

Music labels charge 1¢ for every song streamed.  That’s $10/thousand songs. To support this with advertising, the publishers need to achieve a CPM of $10 with 100% sell-through.  In SAI’s view, that’s not achievable – especially considering additional business expenses and shareholder expectations.

A more realistic CPM (says SAI) is $1.  That translates to 0.1¢ per song.  When compared to the 70¢ the labels currently get at iTunes, the price difference is stunning.  Instead of downloading a song at iTunes, a user would need to stream it 700 times.  For a blockbuster hit that traditionally sells half-a-million tracks on iTunes, that’s 350 million individual streams.

Hmmm. It seems to make sense on a basic level.  

Ahh, but the devil is always in the details.  In this case, God is in the details, too.  Music services like Imeem and MySpace have a lot more options that showing a $1 CPM display ad along with every streamed track.  

 
Brands have traditionally sought music (and paid enormously) for the purpose of aligning themselves with the artist, the statement, the style and the audience.  Consider Michael Jackson as the symbol of Pepsi in the ‘90’s.  Remember the line-up Chevy ad campaigns that were built around songs and personalities like Vince Gill or John Mellancamp.

Second:  One song ≠ one impression.  An average song is  4 minutes long; which give the opportunity for 4 impressions rotating at 1 minute increments.   Also, there’s no rule that says only one ad can be displayed at a time.  All in all, it’s reasonable to expect that one song could generate up to 8 ad impressions.  That changes the SAI calculations dramatically.   

Third:  Consider the context.  Ad revenues from streams only go toward paying the RIAA, but what about business expenses of the publishers themselves?  Well, recall that every song lives on a website/interface that will have additional navigation, search, promotional, and other content pages.  Each of these pages will generate additional ad impressions.  In fact, I would go so far as to assume that for every song played, there could be an additional page viewed either to discover the song, locate the song on site, rate/comment about the song, bookmark the song, manage the song in a library, etc.  

Music licenses are a business expense, and a streaming service only needs for each song to pay for itself.  The profit for the publisher is in the ancillary advertising, merchandising, sponsorships, and other revenue streams.  An astute business will find hundreds of ways to generate additional high-value inventory within their service.

Fourth:  Consider the promotional opportunities.  83% of Americans use the radio to discover new music – that fact isn’t lost on record labels, who pour millions into radio networks in order to promote their upcoming artists.  This is a model which is both lucrative for labels and for radio stations.

Now consider that old-style marketing scenario fortified with 21st century technology.  Not only can labels promote new artists to mass audiences via online streaming services, but now they can benefit from direct response statistics and behavioral targeting.  All of this makes promotional inventory on a widely-used service worth millions of dollars.  

I envision EMI buying a front-page placement for a new artist, targeted to the exact age and gender of the audience, along with matching their exact musical tastes, and then (as icing on the cake) measure the effectiveness of each song and marketing message via A/B testing and click-through statistics.  That’s a $50 CPM if I ever saw one!

I'll leave it at that for now, although I'm sure there are additional ways of making streaming services profitable -- assuming the audience is there.  Of course, in the end, that's the biggest "if".  Will audiences actually respond to and adopt free, on demand streaming services?  As I've said before, I don't believe it has mass appeal.   Time will tell.  But in the meantime, despite the analysis of my friends at SAI, the company should have no problem breaking even on it's 1¢/song licence fee.