Making Applesauce of Apple's Numbers...
I believe David Caulton has been staring into his cracked rearview mirror again (and his little parrot is still on his shoulder). David's view of Apple's posted numbers is skewed a little more with each new post and Pauly continues to praise his data. This is my second reply to David Caultons's iTunes/iPod series. [First Reply]
I like David Caulton. I've enjoyed reading about Zune via Zunester since its creation. From his site, he seems to be a good family man, a loyal employee and a good guy in general. He always seemed level-headed and open-minded about the DAP market. But lately, he's continued to release graphs and statistics based on averages of estimates, approximations of trends and comparisons of apples to... well, different apples. I won't even argue with his fatally flawed iTunes/iPod arguement. My faith in his accuracy and reliability is waning. His steadfastness to deriving specific trends using numbers with such variable periods and reporting dates is pushing the limits of reasonable conclusions. Perhaps he should stick to writing about the things he can't write about - Zune.
Analyzing Caulton's Numbers
Let's start with his most recently collected data, which states that iTunes has sold 2 Billion songs. This is (mostly) true as it was released on Jan 9th 2006 at MacWorld. Regarding this David writes, "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)."
First, let's analyze the 3.8M per day over the period part of David's statement. Best I can tell from his graph, he has the same data that I have of 1.5B songs released at the "It's Showtime" event on Sept 12, 2006. Counting the days until Jan 9, 2006, I figure 119 days. So Apple sold 500M songs in 119 days for an average of 4.2M songs per day. David has 3.8M, I have 4.2M - one of us is 10% off.
Next, David claims that Apple said they sold 5M songs per day at the peak. In reality, Jobs said "we are selling over 5 million songs a day now." I'm sure David just heardwhat he wanted to hear wrong.
Now let's look at what a real "at the peak" figure might reach. According to the Forrester research, iTunes song sales declined in 2006 after the 2005 Christmas boom. Therefore, taking the 2.5M song/day average from Feb to Sept 06 and putting it on a decline means that Apple was selling less that 2.5M songs per day as of Sept 12th and those numbers were still declining. Also, if Forrester is correct, that decline would mean that even fewer iTunes sold from Sept 12 to Dec 24th. Which indicates that Apple was selling less than 2.5M songs/day for 104 days. Now, at less than 2.5M songs/day that means that a maximum of 260M songs were sold in the 104 days running up to Christmas. This also leaves 240M songs to be sold in 15 days after Christmas. This would require Apple selling an average of 12.6M songs/day after Christmas. Of course, did Jobs say Apple reached 2 billion today? Yesterday? A week ago? No, he just said they've passed 2 billion.
But wait, there's more. Jobs also released a graph showing iTunes sales per year. According to Apple, in 2003, 30M; 2004, 191M; 2005, 614M; 2006; 1.2B. Whoa, wait a minute! That means that not only had Apple reached 2 billion before Jan 9th, they reached 2.035 billion on Dec 31st. I won't bore you with the math again, but that means that according to the "sales drop until Christmas" belief, 275M songs were sold in 6 days between Christmas and New Years Day. That's 45M songs per day. I guarantee that peak David speaks of at 5M was Christmas day and was likely over 50M downloads. The downloads clearly outpaced Apple's expectations and crippled iTunes. Don't think for a second Apple hadn't prepared for Christmas morning or so they thought... Do you think Apple wanted new iPod owners first iPod/iTunes experience to be tainted by a "cannot connect to iTunes Store" message?
Now, I am highly doubtful that Apple has ever rounded their sales numbers up. They are a publicly traded company that must answer to shareholders and the practice of rounding sales up and claiming the number as a fact is at the least unethical. So, just like during the MacWorld keynote, Steve likely rounds iTunes sales down. Maybe by a few thousand at the beginning and maybe by millions now days. This time it was a quote of 2 billion when the 2006 annual figure says sales were actually 2.035B nine days earlier. That 35M more songs and 9 days less of a difference make a huge impact on average daily sales as you saw by the 12.6M per day jump to 45M per day.
A Few Graphs
The first graph shows iTunes transaction growth. iTunes tracks receipts of iTunes purchases and numbers them sequentially. As I mentioned in my last post, I have well over a hundred iTunes transactions. Lucky for me I had made a purchase within 2 days of the 1st of every month. This is a graph of those iTunes transaction numbers. I haven't bought anything before April 2004. Also, if anyone can provide transaction numbers for the day before and day after Christmas, I'd be very interested to hear them; of course, they may not matter as people were likely buying $15, $25, $50+ worth of songs per transaction due to gift cards. I know my sister used her $50 gift card in one swipe. Of course, I taught her to add wanted songs to a "Wanted Playlist," so she had a nice queue waiting. Comparing every known song milestone with my nearest iTunes transaction number I found the ratio of songs per transaction is a very steady six. Video and particularly movies may have prevented the songs per transaction from increasing due to the nature of buying one video at a time or maybe six is the magic number. I'd estimate, given that an album is usually over 12 songs and most people go to iTunes with a specific track or album purchase in mind, the average of six songs is correct.
Next is a quick chart just showing the average daily sales per period. I'm not going to take the liberty that David takes and say whether sales rates during each period inclined or declined. The flat lines you see are simply the average daily sales for that period. You can decide if it was an incline or decline based on dates and pre and post-period averages.
All I'm saying is that David is trying to pull VERY specific data, from VERY loose numbers. He averages across Christmas when it proves a point he wants to make and then points out the Christmas spikes when it helps make a different point? His graphs are such wild approximations that they mean very little. A much better indicator is the graph Apple showed, year by year. This accounts for all the seasonality and provides a matching, clear cut off date for both iTunes and iPod of Dec 31st - a detail you can't account for with Apples random milestone releases.
This is a graph of those numbers we know for sure; they're still rounded unless you think Apple sold exactly 1.2B songs for 2006. Jobs released the iTunes numbers at MacWorld and the iPod numbers can be found on Apple quarterly reports. Here's something interesting about this graph related to David's theory of declining iTunes songs/iPod. At the end of 2006, Apple has sold a total of 2,035,000,000 songs. As of Sept 30, 2006, Apple has sold 67,648,000 iPods. In order for the number of iTunes songs/iPod to actually continue declining lower than 19.7, Apple will have to sell over 35.5M iPods this holiday quarter. So it's either Woohoo for stockholders or Boohoo for those that say the ratio of iTunes songs per iPod is declining. No matter what, Apple's doing great. As I said in my last post, iTunes sales are growing faster than iPod sales. Besides, I'm not counting the owners of multiple iPods. How many people got second or third iPods this year? I also said, iTunes/iPod owners is much more relevant than iTunes per iPod. This iPod owner has over 1000 songs and 4 iPods. If you consider that some of the songs were my wife's requests then we each have 550 songs. But according to David, each iPod only sold 250 songs. Sorry David, it doesn't work that way.
Data Used: (iP = iPod, iT = iTunes)
iP - 03/31/2004 - 685 K
iT - 04/28/2003 - iTunes Store Opens
iT - 05/15/2003 - 1 Million
iT - 06/23/2003 - 5 Million
iP - 06/30/2003 - 989 K
iT - 09/08/2003 - 10 Million
iP - 09/30/2003 - 1.325 Million
iT - 12/15/2003 - 25 Million
iT - 12/31/2003 - 30 Million
iP - 12/31/2003 - 2.058 Million
iT - 03/15/2004 - 50 Million
iP - 03/30/2004 - 2.865 Million
iP - 06/30/2004 - 3.725 Million
iT - 07/12/2004 - 100 Million
iP - 09/30/2004 - 5.741 Million
iT - 12/16/2004 - 200 Million
iT - 12/31/2004 - 221 Million
iP - 12/31/2004 - 10.321 Million
iP - 03/31/2005 - 15.632 Million
iP - 06/30/2005 - 21.787 Million
iT - 07/17/2005 - 500 Million
iP - 09/30/2005 - 28.238 Million
iT - 10/25/2005 - 600 Millions
iT - 12/31/2005 - 835 Million
iP - 12/31/2005 - 42.281 Million
iT - 01/10/2006 - 850 Million
iT - 02/07/2006 - 950 Million
iT - 02/23/2006 - 1 Billion
iP - 03/31/2006 - 50.807 Million
iP - 06/30/2006 - 58.918 Million
iT - 09/12/2006 - 1.5 Billion
iP - 09/30/2006 - 67.648 Million
iT - 12/31/2006 - 2.035 Billion
iP - 12/31/2006 - 103.3 Million? Stay tuned.
I like David Caulton. I've enjoyed reading about Zune via Zunester since its creation. From his site, he seems to be a good family man, a loyal employee and a good guy in general. He always seemed level-headed and open-minded about the DAP market. But lately, he's continued to release graphs and statistics based on averages of estimates, approximations of trends and comparisons of apples to... well, different apples. I won't even argue with his fatally flawed iTunes/iPod arguement. My faith in his accuracy and reliability is waning. His steadfastness to deriving specific trends using numbers with such variable periods and reporting dates is pushing the limits of reasonable conclusions. Perhaps he should stick to writing about the things he can't write about - Zune.
Analyzing Caulton's Numbers
Let's start with his most recently collected data, which states that iTunes has sold 2 Billion songs. This is (mostly) true as it was released on Jan 9th 2006 at MacWorld. Regarding this David writes, "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)."
First, let's analyze the 3.8M per day over the period part of David's statement. Best I can tell from his graph, he has the same data that I have of 1.5B songs released at the "It's Showtime" event on Sept 12, 2006. Counting the days until Jan 9, 2006, I figure 119 days. So Apple sold 500M songs in 119 days for an average of 4.2M songs per day. David has 3.8M, I have 4.2M - one of us is 10% off.
Next, David claims that Apple said they sold 5M songs per day at the peak. In reality, Jobs said "we are selling over 5 million songs a day now." I'm sure David just heard
Now let's look at what a real "at the peak" figure might reach. According to the Forrester research, iTunes song sales declined in 2006 after the 2005 Christmas boom. Therefore, taking the 2.5M song/day average from Feb to Sept 06 and putting it on a decline means that Apple was selling less that 2.5M songs per day as of Sept 12th and those numbers were still declining. Also, if Forrester is correct, that decline would mean that even fewer iTunes sold from Sept 12 to Dec 24th. Which indicates that Apple was selling less than 2.5M songs/day for 104 days. Now, at less than 2.5M songs/day that means that a maximum of 260M songs were sold in the 104 days running up to Christmas. This also leaves 240M songs to be sold in 15 days after Christmas. This would require Apple selling an average of 12.6M songs/day after Christmas. Of course, did Jobs say Apple reached 2 billion today? Yesterday? A week ago? No, he just said they've passed 2 billion.
But wait, there's more. Jobs also released a graph showing iTunes sales per year. According to Apple, in 2003, 30M; 2004, 191M; 2005, 614M; 2006; 1.2B. Whoa, wait a minute! That means that not only had Apple reached 2 billion before Jan 9th, they reached 2.035 billion on Dec 31st. I won't bore you with the math again, but that means that according to the "sales drop until Christmas" belief, 275M songs were sold in 6 days between Christmas and New Years Day. That's 45M songs per day. I guarantee that peak David speaks of at 5M was Christmas day and was likely over 50M downloads. The downloads clearly outpaced Apple's expectations and crippled iTunes. Don't think for a second Apple hadn't prepared for Christmas morning or so they thought... Do you think Apple wanted new iPod owners first iPod/iTunes experience to be tainted by a "cannot connect to iTunes Store" message?
Now, I am highly doubtful that Apple has ever rounded their sales numbers up. They are a publicly traded company that must answer to shareholders and the practice of rounding sales up and claiming the number as a fact is at the least unethical. So, just like during the MacWorld keynote, Steve likely rounds iTunes sales down. Maybe by a few thousand at the beginning and maybe by millions now days. This time it was a quote of 2 billion when the 2006 annual figure says sales were actually 2.035B nine days earlier. That 35M more songs and 9 days less of a difference make a huge impact on average daily sales as you saw by the 12.6M per day jump to 45M per day.
A Few Graphs
The first graph shows iTunes transaction growth. iTunes tracks receipts of iTunes purchases and numbers them sequentially. As I mentioned in my last post, I have well over a hundred iTunes transactions. Lucky for me I had made a purchase within 2 days of the 1st of every month. This is a graph of those iTunes transaction numbers. I haven't bought anything before April 2004. Also, if anyone can provide transaction numbers for the day before and day after Christmas, I'd be very interested to hear them; of course, they may not matter as people were likely buying $15, $25, $50+ worth of songs per transaction due to gift cards. I know my sister used her $50 gift card in one swipe. Of course, I taught her to add wanted songs to a "Wanted Playlist," so she had a nice queue waiting. Comparing every known song milestone with my nearest iTunes transaction number I found the ratio of songs per transaction is a very steady six. Video and particularly movies may have prevented the songs per transaction from increasing due to the nature of buying one video at a time or maybe six is the magic number. I'd estimate, given that an album is usually over 12 songs and most people go to iTunes with a specific track or album purchase in mind, the average of six songs is correct.
Next is a quick chart just showing the average daily sales per period. I'm not going to take the liberty that David takes and say whether sales rates during each period inclined or declined. The flat lines you see are simply the average daily sales for that period. You can decide if it was an incline or decline based on dates and pre and post-period averages.
All I'm saying is that David is trying to pull VERY specific data, from VERY loose numbers. He averages across Christmas when it proves a point he wants to make and then points out the Christmas spikes when it helps make a different point? His graphs are such wild approximations that they mean very little. A much better indicator is the graph Apple showed, year by year. This accounts for all the seasonality and provides a matching, clear cut off date for both iTunes and iPod of Dec 31st - a detail you can't account for with Apples random milestone releases.
This is a graph of those numbers we know for sure; they're still rounded unless you think Apple sold exactly 1.2B songs for 2006. Jobs released the iTunes numbers at MacWorld and the iPod numbers can be found on Apple quarterly reports. Here's something interesting about this graph related to David's theory of declining iTunes songs/iPod. At the end of 2006, Apple has sold a total of 2,035,000,000 songs. As of Sept 30, 2006, Apple has sold 67,648,000 iPods. In order for the number of iTunes songs/iPod to actually continue declining lower than 19.7, Apple will have to sell over 35.5M iPods this holiday quarter. So it's either Woohoo for stockholders or Boohoo for those that say the ratio of iTunes songs per iPod is declining. No matter what, Apple's doing great. As I said in my last post, iTunes sales are growing faster than iPod sales. Besides, I'm not counting the owners of multiple iPods. How many people got second or third iPods this year? I also said, iTunes/iPod owners is much more relevant than iTunes per iPod. This iPod owner has over 1000 songs and 4 iPods. If you consider that some of the songs were my wife's requests then we each have 550 songs. But according to David, each iPod only sold 250 songs. Sorry David, it doesn't work that way.
Data Used: (iP = iPod, iT = iTunes)
iP - 03/31/2004 - 685 K
iT - 04/28/2003 - iTunes Store Opens
iT - 05/15/2003 - 1 Million
iT - 06/23/2003 - 5 Million
iP - 06/30/2003 - 989 K
iT - 09/08/2003 - 10 Million
iP - 09/30/2003 - 1.325 Million
iT - 12/15/2003 - 25 Million
iT - 12/31/2003 - 30 Million
iP - 12/31/2003 - 2.058 Million
iT - 03/15/2004 - 50 Million
iP - 03/30/2004 - 2.865 Million
iP - 06/30/2004 - 3.725 Million
iT - 07/12/2004 - 100 Million
iP - 09/30/2004 - 5.741 Million
iT - 12/16/2004 - 200 Million
iT - 12/31/2004 - 221 Million
iP - 12/31/2004 - 10.321 Million
iP - 03/31/2005 - 15.632 Million
iP - 06/30/2005 - 21.787 Million
iT - 07/17/2005 - 500 Million
iP - 09/30/2005 - 28.238 Million
iT - 10/25/2005 - 600 Millions
iT - 12/31/2005 - 835 Million
iP - 12/31/2005 - 42.281 Million
iT - 01/10/2006 - 850 Million
iT - 02/07/2006 - 950 Million
iT - 02/23/2006 - 1 Billion
iP - 03/31/2006 - 50.807 Million
iP - 06/30/2006 - 58.918 Million
iT - 09/12/2006 - 1.5 Billion
iP - 09/30/2006 - 67.648 Million
iT - 12/31/2006 - 2.035 Billion
iP - 12/31/2006 - 103.3 Million? Stay tuned.
Comments and Opinions:
0 showing
Post a Comment