moblog skiing

sent from the backcountry of snoqualmie pass.

another ski shot, with gear.

Music store value chain

Many folks don't understand the byzantine world of music rights and sales, and I thought I'd try to post a one page explanation.

Music happens in several stages:

• A songwriter writes a piece of music.
• They then do a deal with a publisher. The publisher is responsible for collecting publishing royaltees for performances and copies of the music. For example, if a song appears in commercial or movie, or is performed publicly, or a digital copy is made, or sheet music is printed, the publisher and songwriter gets paid. You should think of the publisher as having rights to the composition (or sheet music), but not the actual recordings. To make things confusing, many music labels (a different role described below) also have publishing arms - EMI, Warner, Sony, Universal. But there are many other independent publishing arms. Another source of confusion is the mechanical rights (a payment to the publisher for each and every copy of a composition) vs. the performance rights, both of which are controlled by the publisher. The publishing rights are often a mess, and this part of the system gets really confusing when the world turns digital.
• An performer records or performs the music. Because this is the most public face of the record industry, I'll assume no more need be said here.
• The artist will probably have entered into a recording contract with a Music Label. The Label offers financing and development, along with massive distribution and marketing to the artist in exchange for the right to exclusively market all recordings for a specific period of time. There are a number of independent labels, but the "big four" that sell the vast majority of US music are Warner, EMI, Sony, and Universal. In most cases, the label is paid every time a CD or digital copy is sold.

These three entities are just the surface later of the puzzle, and to prevent your head from exploding I won't go into the Harry Fox Agency, ASCAP/BMI, the variability in all the relationships above, etc....

This whole system seems very painful - and it is. It has evolved over the last century and keeps getting twisted. Remember it started out as being about compositions and printed sheet music sales, then had to evolve to accomodate recording, broadcasting, TV, Movies, and now digital distribution.

Mysterious photo posts - revealed!!

I love the conspiracy theories...but the last couple of posts were just me playing with a new phone (Dash), configuring mms on cingular with it, then getting blogger's newish moblog service running. Pretty painful process, but the results are cool.
On another topic entirely, I'd like to say thanks again to everyone who provided info and feedback on skipping experiences. The tone of the messages has been great and they've all gone to the test team. They are investigating, and for now all I can say is hang in there. We really appreciate your having bought the device and recognize the value of you early adopters...

Planner paluza!

bill!

Fairplay opening up? Tempest in a teapot?

A number of articles have discussed apple "opening up" Fairplay DRM. I think this rumor is based on a program apple's been building up to for a while. If so, afaik it's not a big secret, and it's not as big a deal as some are thinking.

It's about iPod accessories that plug into the bottom of the ipod - docks, car stereos - probably not networked devices (It's unteresting the the article linked above says the EVA8000 supports fairplay content, but the Netgear site specifically implies it doesn't. I'll assume not for now) or other more interesting scenarios. In fact, the devices don't even get to decrypt the content.

To understand the new scheme for these devices, you need to understand the old one. Previously, an ipod dock or car stereo that plugs into the ipod and use a "two wire" solution - one that carries command-and-control (volume up, down, skip, etc...), and one that carries the analog audio. This solution is simple, but has some disadvantages in sound quality, susceptibility to interference, etc.... But it does let the accessory "play" fairplay content (in fact, the ipod is decrypting and playing it).

The new scheme requires that the accessory manufacturer include an "authentication" chip that certifies this device is a good partner, is paid up on licensing, and won't record the audio that it's about the recieve from the iPod. Accessories with the chip can then run a "one wire" solution in which the ipod decrypts the fairplay and then sends the PCM digital audio over to the accessory for playback. This is less secure (theoretically someone could grab the pure digital audio, but in practice probably won't) but it does let the device do the digital-to-analog phase.

I'd assume this "premium" solution costs more, either in a higher license fee or in buying the authentication chip - but I don't know. Apple appears to be using some other additional features (album art, new commands) as additional carrots to get folks to sign up for this.

So, is this opening up fairplay DRM? You be the judge. But it's certainly not the grand opening up that will enable licensing and playback in other software, hardware, and/or competing devices that some are speculating.

AOL exits the music subscription biz - Napster picks up the pieces

AOL/Musicnow is apparently dumping themselves and giving their users to Napster. This follows a similar move by Virgin digital. I'd make the same comments as I did then....

This highlights what's likely to be a trend - there's not really any profit
in running a standalone music service (Apple and Zune are likely to make the
money from hardware sales, not from service fees), so the standalone guys are
left with a real challenge. Expect to see more consolidations, shut-downs, and
outright acquisitions in this space over the next 12 months.

My unanswered iPhone questions

Not that I'd expect answers now, but if folks have authoritative answers I'd love them:

• how will the touch screen experience actually work for usage? Will the alpha keyboard work well for instance? I've never seen it done flawlessly before, but this might be the first time.
• Apple says the phone runs OSX. What exactly does that mean? Is it a mach kernel? Is it running something lighter with an osx shell?
• to what extent is it a new platform? How easy is it to port apps from osx to this? widget?
• How closed? Will apple/cingular allow installation of apps by users? Develpment of apps by random, unlicenced third parties? (appears to be locked per Gartenberg and others)
• Battery life - big unknown until it ships with optimized software, but a key stat.
• Expandible memory?
• The wifi setup/config experience will be interesting. there's lots of room to improve on existing devices there.

If you've got a line on these, or can think of other questions I'm missing, leave a comment!

Keynote thouights

So, I have to stress here this is really just my personal opinion, so for what it's worth....

iTV -er- Apple TV

• Pros: The UI is slick and pretty. The HDD was a surprise, and indicates they don't think they'll always be able to make streaming video work reliably. I played with one today and it's snappy.
• Cons: Will be a challenging sell outside of the enthusiast community. It's a $299 device that lets you watch TV in your living room! Comcast already put one of those in my living room. Comes with the little mac remote control, which feels too small/limited to me. It'll be a great itunes video enthusiast accesory, though.
• My prediction: Apple will make a nice margin and the device will sell 1M units in the first 12 months. It isn't going to get strategically significant numbers of living rooms against xbox, ps3, wii (much less comcast, etc...) But it is a version 1 product, so it'll be interesting to see where they go from here.

the Apple Phone -er- iPhone

• Pros: The most stylish smartphone software ever. Seeing it in action, it's fun to see all the rich UI they have built with the obviously massive processor in this thing. They did a nice job of nailing the basics and going a bit beyond in a few places. Full browser and big screen on a phone is nice.
• Cons: The most expensive phone on the market? Niched by price and formfactor. $499 with a 2 year committment is amazingly high end. I don't have numbers, but that priceband must account for a % of a % of the market. I'm also not a real fan of the big flat PDA phone formfactor, but I know others are.
• Prediction: wow, a 10M goal in 2008? Really? Feels aggressive, but I'd hate to underestimate them. I'd bet on 5M, but again a nice margin and a win for Apple shareholders. That said, I think there's a lot of potential here and Apple will learn fast in the phone space if they can hold the cingular partnership together.

skipping update

A number of comments have mentioned the reports of skipping and unresponsiveness on my part. I'm sorry, and definitely don't intend to bury the topic by any means. For a progress update:

• I took the mails folks sent me, compiled, and forwarded to the test team
• They're investigating and trying to repro to understand the thing(s) that might be causing skipping

When I hear back, I'll post an update.

Macworld Keynote numbers

Just returned from the show. Apple came in right about where I'd expected. My interpretation - the number ipods (and thus itunes customers) continues to grow rapidly.

• They grew the raw rate of sales back up to the rate from last holiday; 3.8M per day over the period, (they said 5M per day at the peak). this supports the omments of many about seasonality in the market nicely.
• they also slightly picked up the sales per ipod; 1.65 songs per month per ipod, up from 1.32 per month per ipod in the last period, but nowhere near the long term average of more than 3.3 per ipod per month.

Boiling it down - songs per ipod.

Man, there's a lot of passion on this issue. The comments on this blog have gotten deeeeeeep into the weeds, deeper than need be. Let me try boiling down to the simplest form.

Between 2/23 and 9/1 of this year, apple sold 500M songs. That during a period when there were well over 60M ipods (say, 50M users accounting for duplicates/replacements). That's 10 songs per user over 190 days or 6.33 months. That comes to 1.58 songs per month per ipod user.

This has several possible source of inaccuracy:

• If there are even fewer iPod users, the number will be larger. for example, if the 60M ipods sold into 30M users, then the number would be 2.78 songs per ipod.
• As songs are bought by users without ipods, the number would be smaller. For example, if 100M of the songs sold to non ipod owners, the number would be 1.33 songs per ipod.

We'll probably find out more tomorrow if Steve announces both iPod and iTunes numbers for December. That may give us another 3 month sample to consider.

Now, I apologise, but I need to get off of the comments for now...Let's revisit after the next data point.

I'm cramming to get packed for my trip to San Francisco tomorrow, but one of the responses was pretty in depth so I wanted to respond. They say:

David believes that the success of the iTunes Store should depend on iPod sales. There is an obvious connection between the two, yet there is no reason one could not be successful without the other. There is not a single reason that songs have to be purchased from iTunes, it's simply an option among dozens of methods for obtaining music. Furthermore, I know over a dozen people who are/were buying music from iTunes, sans iPod. I also know dozens of people with iPods who have never bought a song from iTunes Store - many buy an iPod just to hold their entire CD collection. The device and the store can exist independently.

These are all good points, and true in an individual basis. But they don't escape the basic fact that iTunes sales growth is actually slower than iPod sales growth, and the number of songs bought per iPod is very low (compared to the number of songs bought per CD player in the 1990s). If there were lots of people without ipods buying music, or if ipod users gradually fell in love with itunes and accellerated buying, the opposite would be true and itunes would be growing faster than iPods.

Second, David claims that the iTunes Store had a period in which they sold 3.6M
songs per day. I have yet to find a way to reach this conclusion.

Pretty simple(minded?). Apple announced sales of 850M songs on 1/10/2006 and 950M on 2/7/2006. They thus sold 100M songs in 28 days. Likewise, you can do a similar anlysis for 10/25/2005 (600M) and 1/10/2006 (850M).

His graphs do a good job of talking about itunes songs per ipod, and I think we can draw roughly the same conclusion. Sales per iPod isn't taking off.

The last point I'll address/agree wtih is the point of averaging over variable periods. It's unfortunate Apple won't release daily sales figures (hah) but I think the key points in this discussion are more around long term trends (e.g., overall itunes sales/time/ipod is at best staying constant, or dropping) rather than point-to-point changes. These work nicely in spite of the churn in data points, and in fact you could make them largely go away by doing moving averages.

This post necessarily talks a lot about apple and itunes, and that's bound to bring on the Apple madness again. In a vain effort to avoid that, let me say again that this isn't a criticism of Apple - they've done a good job of parlaying profitable massive device sales into the #1 online music sales channel.

Trends in digital downloads

So, to keep things less inflamatory, let's talk about the trends in the digital download market overall, and why it matters (even if not to Apple) if the digital download market is slowing down.

Take a look at the trends in this market here. All the excitement and press around digital downloads is swept aside by the simple stat that wile digital sold $500M in songs in 2005, the rest of the market (CDs, etc...) sold $11.2B (but those sales are dropping). That's something like 9 Billion tracks per year - if there are 100M music buyer in the US, it's 87 tracks per year, 7 per month!

The key question for the industry is whether digital can grow enough to add to or replace lost physical sales. We get an initial indication here: the growth in digital from 2004 to 2005 (corrected) didn't offset the drop in physical sales.

All the discussion below is around which of these two scenarios will play out.

• Digital takes off! iPods and Zunes sell into as many households as CD players, and their users buy digital music on an ongoing basis. They continue to buy roughly 7 tracks per month.
• Digital stalls. mp3 player owners initially buy 30 tracks in the first 3 months, but then drop off to (my estimate based on iTunes history) 1 track per month or less on average. In the long run, this is a catastrophy for the music biz.

This is why these questions aren't academic. They're academic for Apple, but not for the content industry. And it's why it's important for innovation in content services to continue to make them more attractive to device users over the long term than iTunes (and other current offerings) appears to be.

Update to songs per month per ipod

Whoops - someone pointed out that I missed the 1.5B announcement from September. Here's the updated graph of sales per month per ipod.

What really rocks is that if you compute sales per day, the 1.5B number itself (direct from Steve) confirms the study, that itunes sales dropped from Feb to Sept, as shown on the blue line below.

One comment made me realize I should have been clearer. I definitely think the Forrester data is right; as growth in new ipod users slows, iTunes sales are dropping off and will continue to do so. Apple will get a big burst of sales this holiday as the new millions of users come onboard, and will have a nice announcement next week, but unless they generate long term sustained buying from existing users, it's bound to drop off.

That said, it's not a problem for Apple and doesn't mean Apple or Itunes are in trouble. It just shows that users aren't switching to itunes as their main way of obtaining music.

I'm involved peripherally in a dust-up over iTunes data between John Gruber, Comscore, Rob Enderle, Paul Thurrott, and Forrester research, so I thought I'd comment.

The trouble started with a Forrester research report showing that Apple iTunes music sales had dropped off. I'll give you my data on this in a minute, but first let me point out - even if it's true, it doesn't matter much to Apple. Apple makes essentially no profit off of the iTunes Music Store; the primary value is marketing, and they get that either way. What's more, iTunes has never generated enough sales to create significant "stickiness" or barrier to switching for most users, so there's not much lost there.

Ok, so why isn't this surprising? iTunes sales over time (check here), show a great looking upward trend. A better graph would be songs sold per month, but that has a similar pattern.

The key to understanding this trend is that it's entirely driven by iPod sales. The more ipods there are, the more songs get sold. This is confirmed by looking at songs sold per month per iPod. Here's that graph:

iTunes sales per ipod has been consistent at 2-4 songs per month with a slight downard trend. iPod users aren't buying more songs, there are just more iPod users over time.

There's an even more important question - what's the buying pattern of an ipod user? Do they buy a lot at first and then stop, or do they keep on buying over time? Until recently, it's been hard to tell. With iPod sales growing 3x year over year, at any given time most iPod owners are new owners who would be buying in either case. This year, that's changed. iPod sales are growing MUCH more slowly year over year, and more of the buyers are replacing exiting ipods, so the proportion of users who are new is much lower. If the installed base is sustaining their buying, then you should see continued sales growth. If users stop buying after the first months, then sales should drop.

That's exactly what the Forrester data showed. Sales dropped in the runup to the holiday crunch exactly as we'd predict. I've actually modeled this out and to see the size drop Forrester reports, you'd have to see a pretty dramatic decrease in buying by an average ipod user - as much as from 10 per month to under 1 per month.

So to be clear about my model. The world before last holiday:

• Every single quarter (even post holiday) shows a huge influx of new iPod owners on top of the Installed base (IB) because the iPod market is growing at 3x/yr
• IPod users buy ~30 tracks on average in the first 3 months they own a device. Then they slow down or stop at around 2 songs per quarter.
• Thus, it looks like iTunes is growing continuously because the huge influx of new (iTunes buying) users offsets the IB falling off into non-iTunes buying status.

The world after last holiday:

• IPod sales hit a wall. Now they begin to grow at "only" 1.3x per year.
• Suddenly the influx of new users (buying 30 songs per quarter) can’t keep up with the dropoff in sales from the IB
• ITunes sales level off or even drop a bit.

By the way; nobody should be very surprised; Apple suddenly stopped announcing momentum on itunes in February...hmmm.....

Subscription service finance 101

Several folks have doubted my claim that it's very difficult/impossible for standalone music download subscriptions to make money. I thought I'd get more concrete.

The answer is that it's theoretically possible. It costs around $12/mo/subscriber to run a subscription service. Thus, if you charge $14.99, you're profiting - $3/month yay!

The reality hits in when you consider the twin horses of the subscription apocalypse - TAC and Churn. TAC is total acquisition cost, the amount it costs you to get each new subscriber. This consists of marketing costs, free trial subscription costs, distribution partner bounties, etc... divided by your new additional subscribers. Churn is the % of subscribers you lose each time period (quarter).

For a subscription service to be successful, TAC and/or churn must be very low. An existing music service (I won't name names, but you've heard of them and the numbers are typical) shows the problem. They pay around $150+ per new subscriber in TAC. Thus, at a $3/month profit, they must keep a customer for 4.17 years to break even (and that's without including the time value of the initial $150...).

Fair enough - they're still ok if they keep the users that long. The problem is, this service churns out user at over 40% per quarter! That means on average they keep each user for only 9 months! Thus, they pay $150 to acquire each user and make back $28 in profits. Not a good story.

Now, a non-standalone offering that includes a device is a different story. Your TAC is lower since you acquire customers through (profitable) devices, and your churn is lower since the end to end with a same-vendor device will make the entire thing "stickier".

So, for example, if you don't market the service independently and thus TAC is (say) $30 and churn is a stickier 10% per quarter, you would be keeping customers for 3.75 years/$135 in profit per customer for an overall net profit of $105 (on top of the device profit). A vastly better story.

Next week - Macworld and CES!

Things are heating up again post holiday - for CES Microsoft has a nice presence and lots of announcements, and I suspect Apple will give us something to think about at Macworld. I'll try to post during this period, but will be traveling and doing internal intelligence on both so may be a bit slammed.

For the record, I'm betting on the following for Macworld:

• Mac and iPod momentum announcements (with or without snarky Zune sound bite)
• OS X Leopard preview and launch date announcements
• iLife 07
• iTV announcement and demo, but not available until Feb or March
• Maybe...maybe...the Apple phone.

The latter is most iffy - the noise and rumors are huge, but there's no way to tell the signal from the noise at this point. I'll revisit my accuracy after the event.

DVD Decrypter - shut down

I'm sure nobody reading this post knows anything about it, but I recently discoverd that DVD Decrypter has been officially shut down by Macrovision. When I went to investigate, I found this story. There are several amazing things about this:

• It took amazingly long - like 2-3 years - to shut down this blatant bit of questionable software. Every time I saw it was still up in very public venues, I shook my head in wonder.
• Apparently Macrovision shut it down in part by buying the software and then issuing cease-and-decists against hosting web sites.

Of course, the Google Adwords banner on the right side of the story hints at the future of this space...

Music sales in the clear?

Obviously, this would be great. Music sales in clear formats would probably be the best thing for everyone.

My prediction is that things aren't quite bad enough for the music industry to do this in 07; the labels will investigate the option thoroughly, maybe do more very limited promotional trials, run up to the cliff, and then back off at the last minute and do somethign more conservative.

I suspect in the long run it's inevitable, though.

Virgin Digital Music Shuts down gracefully

Paul Thurrott over at the Internet Nexus reports on the apparent shut-down of Virgin Music downloads in the US. Apparently they've struck a deal to hand off their music subscription customers to Napster.

This highlights what's likely to be a trend - there's not really any profit in running a standalone music service (Apple and Zune are likely to make the money from hardware sales, not from service fees), so the standalone guys are left with a real challenge. Expect to see more consolidations, shut-downs, and outright acquisitions in this space over the next 12 months.