iTunes sales are NOT plummeting! Press credibility, on the other hand . . .

[Josh] What an interesting couple of days it's been. What follows is a case study in how information -- and misinformation -- spreads on the Net.

We put out a simple little report about iPods and iTunes based on credit card transactions and publicly stated Apple data. And for those who aren't Forrester clients, I blogged the highlighs. In case you are wondering, we ran the report by Apple, and they declined to comment.

Since then:

- The New York Times ran a little fairly balanced pieced on the research. This got us on the media's radar screen. Then . . .

- A UK outfit called The Register and Bloomberg decided to dive in and highlight one finding of the report -- that iTunes sales had dropped in the first six months of this year. We got treated to wonderful headlines about iTunes sales "collapsing" and "dropping" and "plummeting" and so on. Now for the record, iTunes sales are not collapsing. Our credit card transaction data shows a real drop between the January post-holiday peak and the rest of the year, but with the number of transactions we counted it's simply not possible to draw this conclusion . . . as we pointed out in the report. But that point was just too subtle to get into these articles.

- Apple's stock actually did plummet -- 3%. I started getting calls from hedge fund managers. Apple's spokesman called and, although they refuse to go on the record with any facts, they're clearly upset. And I also heard from the Chicago Tribune, San Francisco Chronicle, LA Times, Financial Times, Toronto Globe and Mail, thestreet.com, etc. At this point I was trying to get people off the "65% drop" idea and onto some of the more interesting ideas in the report, with mixed success.

Now, you can't unring the bell. But I will try to focus you on the truth here, which is this: iTunes sales are leveling off, the Journal did an article about it last Friday with data from Soundscan. Apple is not in trouble -- it makes its money mostly from iPods, and iTunes is just a way to make that experience better. It's the music industry that has to worry, since the $1 billion a year or so from iTunes, globally, doesn't nearly make up for even the drop in CD sales in the US, which are now down $2.5 billion from where they were.

Finally, a word for Apple. Apple is extremely stingy with information about their business and public comment. Their unwillingness to comment on the record or off about anything they're working on or any industry results beyond the basic statistics fuels speculation, pro and con, from their supporters and detractors. In the research business we like facts -- and every other technology company is more open with them. So maybe it's time for Apple to share a bit more. When the real bad news hits -- and it's inevitable, no company gets everything right -- that openness would pay off.

I'd really like your comments ... interested in hearing from you.

Technorati tags: Bernoff, Forrester, Apple, iTunes

Comments

[GadgetGav]

It's a shame that this blog entry is not getting the same press that the rest of the (wrong) story has. I even heard the 'sales down 65%' line on NPR this morning.

How did you come up with the 65% figure..? It must have been in your report somewhere. It's hard to see how you can try to do a worthwhile analysis of iTunes sales when Apple doesn't release data. How did you account for iTunes gift cards and other forms of payment, like codes from Coinstar machines..?

Press credibility may be suspect, but you provided fuel for the fire.

[Futurehouse]

I was actually interested enough in the "story behind the story" to spend 10 minutes searching on-line to find out your name and blog.

On the face of it, I was ready to assume that the Forrester report that you were an author of was at fault. Your reasoned blog entry makes me wonder where the truth lies.

Our news media is broken, and I am no longer sure that it can be fixed, even if we wanted to. We thrive on the blaring headline, and the black and white (excuse the pun) interpretation of all stories.

Anyway, thank you for providing some insight into this story.

[timk]

Thanks for the clarification, and I had suspected that there was a lot of misinterpretation going on since the # of iTunes per iPod has actually increased from ~20 to ~22... I very much understand the dropoff as well, but I would like some clarification: is the drop just measured from the Jan-Feb time period to the next month? To the end of your data (I'm hearing June)? Since you have 27 months of data, why wouldn't monthly or any period of growth be measured year-over-year? Did you compare the decline from Jan-Feb of this year to June versus the same time period froma year previous.

Since last year was a big growth year and you seem keenly aware of the seasonality of iTune purchases )post-holiday rather than pre-holiday), it seems like this could have been better contextualized, warned against.

[Remy Fiorentino]

Remy Fiorentino from Forrester here, I helped Josh with the analysis for this report.

TIMK - The 65% decline in revenue was between January 2006 and June 2006. This decline is statistically significant, but is still based on a sample of 181 iTunes buyers. When we compare this to 2005 data, we see that revenues declined 39% from January 2005 to June 2005. Is 2006 worse than 2005? It's hard to say. However, the January 2005 revenue was a bit higher than the January 2006 revenue, and the June 2005 revenue was more than twice the June 2006 revenue. Our data suggests that the iTunes growth has slowed, though neither Josh nor I would say sales have "collapsed".

[PV]

A few questions:
Is 20 (or 23) songs/iPod a failure?

Have you considered the impact of the multiple iPod family/friend network? I fit the average: 20 songs purchased through iTMS. My iPod shows a different story as many of my songs were purchased by other household members and shared within the family. In other words, there's a difference between songs purchased by the owner and the number of purchased songs on an iPod. Although this doesn't affect sales data, it changes the dynamic of legal iPod usage.

You claim that sales have recently moved to 23 songs/iPod sold. Isn't that growth?

[GadgetGav]

Having chided Apple for not releasing more information and having started the story that lead to a 3% decline in the company's stock, it would be nice if Forrester could find some way to release this report into the public domain rather than making us pay $250 to see the real figures and analysis...
I realize a lot of work went into this, but I'd assume it has a limited life to a very limited market. Once your initial sales have dried up, will it be released for free..?

[timk]

Thanks, Remy.

That's what I thought. Who reports on any such random six month allotment? (I don't want to be overly critical, I appreciate the response and the discussion.) But that, right there, is why the finger should be pointed at you guys as well as the media. Why did you neglect year over year entirely? Piper Jaffrey seems to be spanking you on this, and Josh's attempts to claim the report is accurate but the media got it wrong seems strongly diluted by this significant lacking.

"However, the January 2005 revenue was a bit higher than the January 2006 revenue"

This doesn't seem plausible; do you have that correct? Looking at your own tables (http://blogs.forrester.com/.shared/image.html?/photos/uncategorized/itunes_and_ipods_1.jpg) and knowing that prior to last year we were talking about increments of one or two million per month rather than millions per week and that last year was Apple's biggest year for holiday sales of both products, I think you have that backwards. There certainly was more revenue in January 2006 than January 2005. No? Certainly if it was lower, someone within the financial community would have picked up on it in the 11 months since...

From there, it gets difficult to question whether June '05 was much better than this year. However, I would repeat the question: did you see June as the low point for sales in the 27 months of data? If not, which month? If there was not a specific month per year, what fluctuations did you see? And, additionally, rather than saying we cannot confirm a dip because of potential seasonality, could you have said, there is a strong indication that there is seasonality (if you are seeing an uptick around the holiday season)?

Also, previously, I hadn't suspected your sample pool. But now I know we are talking about 181 purchasers out of 2,791 purchases (or 15.4 purchases per purchaser) over 27 months to represent a market of 60 plus units with 1.5 billion tracks... and I'm getting concerned.

Also, can you explain how if the number of iTunes per iPod (despite the poor waulity of such a metric for obvious reasons) has continued to grow (from below 20 to above 22) according to your own numbers, how is the market slowing? I simply cannot understand that. It seems to me that based on your own metrics and numbers, the iTunes market is growing faster than the iPod market. Almost negligibly, yes, but not slowing at all.

Thanks again. To both you and Josh.

[timk]

Also, Remy, based on this comment: "However, the January 2005 revenue was a bit higher than the January 2006 revenue, and the June 2005 revenue was more than twice the June 2006 revenue."... Clearly you are jsut portraying data for two months and there could be significant data gaps but... January is most certainly the biggest month.... According to you, the revenue declined in '06 from '05 for their biggest month and for )possbily) their slowest month revenue declined from '05 to '06 by HALF? Wouldn't that suggest that not only has growth slowed, it's disappeared and actually declined?

Rather than clarified, I think you've confused the situation a bit more.

[mark]

You claimed to have used Apple's public data, but that data from Apple's press releases only shows that there is a little bit of a dropoff. Did you try to reconcile it with your credit card or Soundscan data?

From 2/23/06 to 10/26/06 (or somewhere up to 20 days before that), iTS sold .5 billion songs, for a rate of 2.04m per day. From 7/18/05 to 2/23/06, iTS sold .5 billion songs, for a rate of 2.27m per day. From 1/24/05 to 7/18/05, iTS sold .25 billion songs, for a rate of 1.43m per day.

So big increase from 05 to 06 but small dropoff across the last two .5 billion data points, which could easily be accounted for by seasonality.

I don't have your report but you could have easily made this simple math clear for all.

[mat]

Josh -- I was totally taken aback when I first saw this yesterday via Engadget and wrote a long post (http://mac.honan.net/2006/12/itunes-collapse-orly.php) trying to get to the bottom of the "collapse." Though I had only seen your previous post on the subject, it looked to me like those numbers were used without the proper context--especially the seasonal sales information.

But I have a question. Looking at the RIAA's numbers (http://www.riaa.com/news/newsletter/101206.asp) it would seem that digital sales are still trending way up:

"On the digital side, revenues from various formats, including downloads, kiosks and digital videos, grew by 86.6 percent in the first half of 2006, with digital formats' share of overall industry value growing to 18 percent in the first part of 2006 -- up from 9 percent in the first half of 2005."

Did your findings run counter to this?

[Scott Loftesness]

Josh,

As a payments industry professional, I just have to ask where you got your "credit card transaction data" that you're saying was the basis for conducting this analysis?

Scott Loftesness, Glenbrook Partners

[philip]

"However, the January 2005 revenue was a bit higher than the January 2006 revenue, and the June 2005 revenue was more than twice the June 2006 revenue."

I'm sorry, but that does not make sense. Can you clarify the statement?

You are saying that revenue for January 2006 was slightly less than January 2005, and revenues for June 2006 were significantly less than June 2005? That just doesn't wash: from 2005-2006 Apple saw the store increase its sales rate by 2.5x. How could you possibly be the only people to observe them making LESS money? What you are saying would not mean growth has slowed (as you have reasserted), but would actually mean sales have declined.

I assume you misspoke so I would appreciate some clarification.

[Gunnar]

Big symbols make big targets... Especially ones that are as popular as Apple, Inc.

[Eric Pinczower]

Obviously your remarks were taken out of context - no one could possibly imagine than itunes sales had dropped 65% - what suprised me was that so many "investors" bailed and the stock dropped so much - by the way thank you - I bought calls at the nadir yesterday and sold them today....... sweet.

[Gospel Moody]

The two graphs I've seen show no drop off anywhere that I can see. Apple sold a half billion songs up until last year and another billion since (if that info is correct). So again, where is the drop off? Maybe you are on the up-and-up on this but after the shenanigans of the MSM in the past few years, I'd be a fool to believe. Emphasising a seasonal trend to make it look as if it is the complete thing, eh? Just so you could say "Gotcha"? Your explanation doesn't wash with me. But, as I say, maybe you are stand-up people and I just have it wrong.

[Dave]

The only one who's credibility is tarnished is Forresters.

Bad data. Poor analysis.

Any other datamongers to get info from?

their blog is trying to spin the earth in the other direction

[Paul]

You guys realize this whole firestorm was actually created by Andrew Orlowski at the Register, who has a history of writing iTunes-bashing articles where he magnifies some tiny thing into a world-ending Godzilla problem.

Orlowski, as usual, took a rather mild report with uncertain conclusions and then ran with it because, let's face it, he personally wants to see the destruction of iTunes ever since he predicted in a ranting piece years ago how the "intolerable" FairPlay DRM would doom iTunes. That hasn't happened and Orlowski is still seething.

Orlowski used you guys, and if you want to correct the record, you need to stand up and call the Register out.

[Nick Carr]

Josh,

Regarding your statement, "Our credit card transaction data shows a real drop between the January post-holiday peak and the rest of the year, but with the number of transactions we counted it's simply not possible to draw this conclusion . . . as we pointed out in the report."

Can you specify where in the report you "pointed out" any cautionary statistical caveat regarding "the number of transactions we counted"? I've gone through the report pretty carefully and can't find any warning about the sample size. In your Methodology note, you simply specify, without comment, the iTunes sample size as being 181 households, after describing the overall panel as being "nationally representative."

If you don't have confidence in the 181 household sample size, do you, in retrospect, think that you might have been clearer about that lack of confidence in the report?

Nick

[Terri Hunt]

Your comment that these sorts of things wouldn't happen if Apple was more forthcoming with facts is nothing more than blatant buck-passing. It also contains more than a hint of blackmail. You managed to garner some media attention with your sensationalistic and shoddy "research", but your reputation has surely been harmed in the process and your failure to step up and accept the blame, rather than blaming others, merely makes the situation worse.

[Matthew Rigdon]

Welcome to the new age, where everything you say will be taken out of context!
Music is something of a weird beast to analyze, isn't it? I mean, you can't quantify taste. Frankly, this year doesn't seem to have resulted in any particularly noteworthy music releases. In fact, the only new song I can think of this year that's had any impression on me is that stupid John Mellencamp song that's always playing in the Chevy commercials (to be fair, the song isn't all that bad, it's pretty standard Mellencamp, but I hear it at least twenty times during every football game. Hey, when I was a kid I loved devilled eggs until I ate a whole plate full and puked. Now I can't stand them. Ad guys might want to look into that effect.)
It seems that, since the downloading game is built on singles, if there isn't a lot of really strong tracks out there, downloads will suffer. If you want an album, you'll go buy the CD.
And who knows? Maybe the availability of individual tracks for sale online has pushed the labels to put together better ALBUMS to get us to go buy CDs. Let's face it, most releases are filled with lots of garbage to fill in the holes between the two or three songs that will go out to Clear Channel.

[Thomas]

This is just another example of how the media sensationalizes facts to pursue readership and in the process becomes a distortion machine. Is there any doubt why they have serious credibility issues with the public?

[Jon]

Sales will "plumet" again this year in Jan feb. Why?
Because people get iTunes gift cards that they spend over the following months.
Your "Research" does not account for the fact that many (maybe most) people have multiple iPods - they don't buy songs for each iPod. More and more people are on their 2nd or 3rd iPod.
You should refer to:
feed://blackfriarsinc.com/blog/atom.xml
For a more interesting take on this.
I think we all should have learned by now not to take this kind of half thought through research too seriously - except as it affects the way people are going to respond to it. So - you created a good buying opportunity with your foolishness. Thanks!

[Steve]

From the graph from the original post looks like the # of songs sold on iTunes is still outpacing the # of iPods sold and the iPod line is shallowing at a greater pace than the iTunes track line. Keeping in mind that iPods are available in a greater number of areas than iTunes service and that the number of iPods actually in service is much less than the total number sold, this looks pretty good for iTunes. The actual sales figures from this curve suggest that there would have been about 650 million sales in 2005, compared to about 800 million in 2006 (if you continue the curve). That's about a 23% increase in total sales over last year. I'm not sure how many other regions they offered iTunes in during 2006 above what they offered in 2005, but I expect that it is less than the 23% increased total sales.
I'd have to take a look at the entire report to be able to say much about the statistical merit of it. The 181 buyers mentioned which implied a decline in revenue of 65% could mean a lot of things but I can't really tell without seeing all the numbers and how this sample is biased. A lot depends on what their exposure to digital music has been over the past couple years. I think when people move to digital music they'll initially buy a lot of music, but will then slow down and go back to their regular purchasing habits. Depending on when those 181 people switched to more digital purchases, you could definitely see a decrease in their purchases. What's more important is how many other people will switch. That's what actually fuels the digital music market, not what people who have already switched to it are doing. Apple is more interested in the growth of their sales, which from the graph in the original post, seems to be going quite well. Once they have had people switch to digital, their job is pretty much done. As the digital market grows to 20 or 30 percent of music sales, the iTunes market will grow at a similar rate as they take on additional customers who will initially purchase a lot of music which can then taper off. It's the increased customers that will help them more than the rates of music purchase from old customers. It's up to musicians and labels to keep people interested in new music, not resalers.

[Muckraker]

As a journalist for the major media who has covered everything from banking scandles ot the music industry, and has had to do extremely difficult forensic accounting, I just cannot believe that so many news outlets ran so many stories that didn't look into the numbers. While that is certainly their fault, I also think that you all are almost criminally negligent in creating a report based on such low numbers of iTunes buyers. 181 buyers! That's nothing. Even 2700 credit cards is nothing. For one thing, do you know that the buyers didn't use other credit cards? For if you are using seasoned buyers, they will have filled up on their favorite old songs first, and bought fewer new ones. Like people on their third and fourth visit to the buffet, they will put smaller and smaller amounts on their plate. New iPod adapters however, will be hungrier, go to the site and fill up. On top of everything, it doesn't even matter how many songs iTunes is selling if Apple earns so little money off of them. If you earn 0% on a million downloads or 0% on a billion, it's still zero. That's not where they make their money.

[Steve Philips]

REFRESHING!!!!!
So good to see the source of mis-used information clarify and point back at the "Journalists" who were at fault.

I'm beginning to think that the only "cure" is to have a "licensing" criteria for "Journalists". A code of ethics, like doctors, psychologists etc. (Lawers intentionally left out of any discussion of ethics!)
In order to represent themselves as a "Journalist" they have to pass tests of their understanding of laws and also submit to a licensing board. If they get so many complaints of violations - plagiarism, negligent distortion of data or facts, hiding behind shield laws after soliciting illegal information etc. then they would be "decertified" and no longer to call themselves a "Journalist" and no longer eligible for any legal protections, badges of entry etc.)
Yeah, THAT would show them!!! :)
Like dreaming of ethical politicians and lawyers.