Does Anyone Remember Napster?

I still perfectly remember the very first mp3 player I ever purchased. I entered college with a 500mb hard drive and quickly ran out of room. Winamp and Napster were conspiring to fill my hard drive with music; both from my own music collection and from the collections of other college students everywhere. I survived the first three years of college with HDD upgrades and some library flushing, but that practice was getting tired and expensive. Plus, it lacked the mobility required to truly benefit from digital audio compression that reduced file size by 90% to 95%. Toward the end of my junior year (2001) came salvation: the mp3 discman.

Physical storage based mp3 players were emerging at the time, but were severely lacking in form factors (we lovingly called my roommate’s player the “Archos Brick”) and/or storage. Topping out at about 100mb, these early, compact players had limited capacity of 20 to 25 songs; only marginally superior to the incumbent Sony Discman. At this point, 2.5” HDDs had broken well through the gigabyte barrier, but very few of us (even techie/nerdie engineering students) could foresee the rapid density growth and price compression that would allow dedicated, large capacity portable media devices. This was especially true in the face of CD-R media costs of mere pennies. The mp3 discman had the promise of portability (it was after all, just a discman) and storage capacity. Each blank CD could be loaded with 700mb, or roughly 175 songs). The navigation was wretched: imagine trying to jump to song 50 on a 60 song playlist using only track-skip buttons. As I exited my junior year in June of 2001, I felt my tech savviness was intact and my music needs were thoroughly satisfied. Four months later, Apple introduced the 5gb first generation iPod, destroyed every existing paradigm and never looked back.

Why am I reminiscing about severely short sighted technology hiccups on the road to ubiquitous music in our pockets? Tech blogs have been buzzing for the last few weeks (really since Apple started building it’s datacenter in NC) about streaming/cloud music services from Apple, Google and, to a lesser extent, Amazon. While it’s unclear exactly what these services will offer (pure digital locker, unlimited streaming, radio-like qualities, social features, etc.), it seems clear that they are intended to compete with Pandora, Spotify (imminent US launch), legacy business models (iTunes EST and Amazon Unbox) and with each other. In other words, the marketplace is about to get extremely clouded…err crowded. Ok, horrible pun.

I understand the desire for Google to get in the music streaming business. To begin with, I am sure that the number of Android smartphone users dissatisfied with the media management options is not insignificant. One friend, a Droid Incredible user, described to me the almost MS-DOS like way he puts his music (historically managed in iTunes to go with his iPod) onto his phone. To him, it almost renders the phone worthless for content consumption. Considering the rapidly changing way in which people are consuming content, this has to unnerve Google (and their handset partners). If there was ever a way to potentially break this cycle, it’s by elevating consumption above the legacy connect-to-synch process. I’m not sure this will work, since wireless synching is something Apple can basically turn on with a switch, but still probably Google’s best shot.

At the same time, content discovery and consumption is moving away from search and [back] toward social/sharing. Actually, to be honest, I’m not sure how much content discovery was ever done by search, other than specific targeted searches. I think hundreds of thousands of college students have built music libraries by “searching” their favorite peer-to-peer (P2P) service. Music has always been be consumed socially. Before the mix tape/CD, people happily had their friends over to listen to the latest vinyl LPs and went to concerts together. Digitization, coupled with broadband, transformed a mix-tape into a mix-library and made sharing (legal and otherwise) the standard way to distribute and acquire new music. I don’t want to delve too much into the impact on the music production/recording industry. That’s an entirely different, though related, series of posts. I will simply say that the network externalities resulting from the digital socializiation of music aren’t all negative. Monetization is still in the early phase, which is the somewhat convoluted focus of this post. The push from multiple service providers to make music access ubiquitous should further lower the barriers to engagement, sharing and (eventually) growth in monetization.

What does this mean for Google? First, I think it’s going to have to link its service up to existing social networks. Orkut, Buzz and other Google social networks have failed. I’m surprised we haven’t seen anything aggressively built into Android. Letting users connect via existing social networks will only bolster engagement with their new service. As long as the interface is intuitive, this will solve the human sharing element. Google can then focus some engineering horsepower on developing a strong algorithmic recommendation engine. Strong recommendation engines have shown to successfully reduce subscriber churn and increase user satisfaction (see: Netflix). This will also give Google the chance to incorporate its best in class ad delivery platform (text and display depending on the content consumption platform), potentially holding down the sticker price for a paid, premium service.

Last thoughts on Google: music might be a chance for Google to forge a strong working relationship with the content owner from the outset. Youtube grew rapidly with by casually looking the other way (ie exploiting) pirated content. Google upset a large number of publishers when it launched its Google Books search initiative. With a music streaming service, they can launch promising music publishers a share of a dual revenue stream and access to over 100mm Android users (many of whom, as mentioned, are looking for a better music option).

Then there is Apple; the $320b pound gorilla. Rumors about an Apple cloud based streaming service have been around forever and pick up a little bit of momentum whenever Apple does anything music related (I bet a Google search history would look sinusoidal). First there was the announcement of their huge North Carolina data center. Next was the purchase of music streaming service LaLa. Then the launch of Ping, Apple’s music social network. More recently, Apple has quietly signed a number of deals with large recording labels. Finally, Apple recently purchase iCloud.com (which had changed the name of its own service to “CloudMe”). Clearly Apple has had its sights set on putting content into the cloud. But to what end?

I am a firm believer that anything Apple does, content-related, is a little bit bi-polar. When it comes to Apple Inc. profitability, content is purely a commodity used to help accelerate hardware sales growth and reduce handset churn. They often say this on earnings calls and in corporate presentations, but I think investors are too clever by half and wary to take Cupertino at their word. It’s pretty straightforward: build a music library in iTunes with an iPod and you will be more likely to replace that iPod with another iPod (or upgrade to an iPod Touch or iPhone). Get comfortable with an iOS handset and you are likely to buy an iOS tablet (or soon an iOS/OSX hybrid PC). Apple has talked about the product halo affect, which helped boost Mac sales from 2% to 10% North American market share over the last 8 years. As I mentioned, even my friend who wanted an Android handset continues to maintain his music library in iTunes and use an iPod. An unlimited music streaming service, coupled with the iTunes and Ping platform, fits snuggly in the existing Apple model.

One model would have Apple “authorize” cloud based streaming of any mp3 previously purchased or ripped from a CD. This authenticated model would do little to disrupt iTunes’ current revenue model and would allow the consumer to virtually synch their i-Device to their music library wirelessly. It could also potentially free up some space locally (though this, I think, is getting to be less and less of an issue). It would be a unique value offering from Apple, allowing the customer to benefit from large, existing iTunes libraries. It would also put iTunes music re-download rights in-line with iTunes [purchased] movie and app rights. The other option is a basic streaming service similar to Rhapsody or Napster/BMG. Here, the consumer pays some subscription fee for unlimited access to whatever music Apple has licensed for the service. Spotify currently runs a sort of hybrid version of these two business models, including tiered pricing for limited usage and advertisements. This is similar to what I expect from Google. The first option is less disruptive to Apple’s incumbent music distribution and management platform. The second is likely to devalue music further, maintain a hold on iOS users and drive continued adoption of Apple hardware. While it might depress iTunes related revenues in the short term, these are low-to-zero margin revenues. The service will continue to shift the “value” from the content to the Apple device on which it’s being consumed.

Here’s what I find strange about this shift in value capture and its implication for the music industry: I really believe that Apple employees (from Steve Jobs all the way down) are hugely passionate music fans. You can hear it in the song selection at Apple product release events and in their commercials. It’s obvious in the C-level excitement about and in the high price they paid for Beatles exclusivity (remember, iTunes music sales aren’t profitable). There are obvious signs of music appreciation. At the same time, Apple is widely blamed in the music industry for demanding 99c song downloads and thus accelerating the decline in album sales. I firmly believe the opposite is true: without the convenience and low price of iTunes mp3 tracks, no one under the age of 35 would have purchased any music in the last ten years. I had a disagreement with a media executive who argued that. $1.99 song downloads would have done just as well, but maintained the value of the content. I don’t think that’s credible in the context of history. At the time, a consumer could buy a 12 to 15 track CD (something tangible) for $15. They were unlikely to pay $25 to $30 for the equivalent digital product. Even if they only wanted two or three album tracks, the perceived price/value of the offering was very low compared to stealing. The digitization of music, the introduction of the mp3 and broadband on college campuses combined to kill album sales. The justice system may have killed Napster, but it was Apple that got young consumers to buy music again.

It’s still extremely easy to pirate music (heck, it’s getting pretty easy to pirate long form video content in HD). Digital music sales growth has disappeared. The music business is now desperate for distribution innovation. We know we are going to see offerings from Apple and Google soon, though it’s unclear exactly how “new” they will be. These offerings will have to compete with some interesting services already on the market.

Spotify is currently a relatively unknown entity in the US, but has over 10mm subs in Europe (of which 1mm are paying subscribers). The desktop application essentially combines iTunes (the ability to buy tracks a la carte), the cloud (mobile access to any songs verified in my home music library), last.fm (quasi-custom digital radio) and, most recently, Napster. P2P functionality is currently somewhat limited, but it introduces intuitive social functionality into a music service. If I want my friend to hear a specific song, I simply drag-and-drop it onto their name and it shows up in their Spotify inbox (I can include a short note as well). This is key functionality that Apple’s Ping lacks; the Ping feed is too passive. Spotify, like Google and Apple, is in the process of negotiating licenses with record labels in the US. I was fortunate to receive an invite for the US trial and think extremely highly of the service.

The other main incumbent, in my mind, is Pandora. Pandora exists only as a streaming service. Their underlying algorithm is the most impressive recommendation engine out there. The service has mapped out a “music genome,” identifying and tagging songs based on underlying characteristics like melody, harmony and rhythm, orchestration, lyrics and vocal harmony (among others). Then, after the user identifies bands, singers, albums and songs he enjoys, it constructs a custom radio station to match your tastes using similar underlying musical traits. The algorithm constantly tweaks the songs it plays based on a simple thumbs up/thumbs down rating system, recursively building a music stream perfectly suited to the individual consumer’s taste (or moods, as Pandora allows multiple “stations”). Pandora has seemingly settled some royalty issues that threatened the company’s future and has moved forward with both an ad-supported and subscription revenue models. Because of its unique underlying technology and playback model (it really is a radio service), Pandora has a large head-start coming competition among streaming music services. It’s extremely important that they continue to innovate. Apple ($65b in cash) and Google ($31b in net cash) have extremely deep pockets and can spend a lot of time and money driving share gains. I’ve already expressed my strong belief that fingerprinting (in this case audio) offers content distributors an easy path to enriching the user experience. It’s time for Pandora to buy Shazaam or SoundHound or to get connected with the open source MusicBrainz project and allow users to tag songs as direct input into user station algorithms. These services/applications need to be seamless to keep users engaged. Otherwise, they will go elsewhere.

In truth, I’m not sure any of these will be “the,” or even “an,” answer for the music industry. I am glad to see the major labels cooperating this time around. Last time around, they went after P2P networks and the “Archos Brick” owners without a viable alternative offering. In that battle, technology won and music lost. Hopefully, this time the consumer can win.

Update (5/4/11 at 9:00am): Looks like Spotify is actually going in the opposite direction and offering a download service.

(Post title a crude reference to “Almost Famous,” which in itself, is a reference to Led Zeppelin live performances)