What if we count the iPad as a Computer…
Remember this graph from Fortune?
Source: Fortune
I found it to be a bit odd.
Reasons:
- the iPad hasn’t only been on sale in the third quarter of 2011 – why didn’t Deutsche Bank update Q3/2010 to Q2/2011 as well?
- the dropoffs of the other manufacturers are too steep – I know the Gartner numbers for Q3 from memory reading tech blogs
So I went and did the legwork the Deutsche Bank analyst obviously didn’t want to do. Six hours later this is the result.
(global PC market share with iPad)
I took my numbers from Gartner and Apple PR press releases. All of them are available online, you can find a list of the sources below.
Something must have gone horribly horribly wrong over at Deutsche Bank because I just don’t see THAT much of an effect. I don’t know what Chris Whitmore did but there’s just something wrong with his math. Global PC sales were 92 Million in Q3/2011 including 5 million Macs sold by Apple. They also sold 11 Million iPads which brings the total number up to 103 million „PCs“. If you look at the Deutsche Bank numbers HP has a market share of 17% (Gartner gives it as 17.7%). To make them drop off to under 15% as the Deutsche Bank graphs shows Apple would have had to sell over 18 million iPads which they simply didn’t.
If you ask me the Deutsche Bank analyst simply added the Macs (5 million) to the iPads in his last column, making it 16 million „Apple PCs). I can’t think of another explanation. Problem is that Macs are already included in „Global PC sales“.
These dropoffs are just too steep without an error.
As a check this is what my graph looks like NOT counting the iPad as a computer:
(global PC sales – market share without iPad)
Which looks pretty accurate compared to the graph from Deutsche Bank.
Sources:
Gartner:
Q1/2008 http://www.gartner.com/it/page.jsp?id=939015 (Q1/2009 report with final numbers)
Q2/2008 http://www.gartner.com/it/page.jsp?id=1076912 (Q2/2009 report with final numbers)
Q3/2008 http://www.gartner.com/it/page.jsp?id=1207613 (Q3/2009 report with final numbers)
Q4/2008 http://www.gartner.com/it/page.jsp?id=1279215 (Q4/2009 report with final numbers)
Q1/2009 http://www.gartner.com/it/page.jsp?id=1353330 (Q1/2010 report with final numbers)
Q2/2009 http://www.gartner.com/it/page.jsp?id=1401136 (Q2/2010 report with final numbers)
Q3/2009 http://www.gartner.com/it/page.jsp?id=1451742 (Q3/2010 report with final numbers)
Q4/2009 http://www.gartner.com/it/page.jsp?id=1519417 (Q4/2010 report with final numbers)
Q1/2010 taken from Q1/2011 (see below) – preliminary numbers: http://www.gartner.com/it/page.jsp?id=1353330
Q2/2010 taken from Q2/2011 report (see below) – preliminary numbers: http://www.gartner.com/it/page.jsp?id=1401136
Q3/2010 taken from Q3/2011 report (see below) – preliminary numbers: http://www.gartner.com/it/page.jsp?id=1451742
Q4/2010 http://www.gartner.com/it/page.jsp?id=1519417 (Preliminary)
Q1/2011 http://www.gartner.com/it/page.jsp?id=1632414 (Preliminary)
Q2/2011 http://www.gartner.com/it/page.jsp?id=1744216 (Preliminary)
Q3/2011 http://www.gartner.com/it/page.jsp?id=1821731 (Preliminary)
all Gartner sales figures were taken from the following year’s report because then they were finalized.
Only the last four quarters have preliminary numbers. But looking at the data from the years before the changes were in the tens to hundreds of thousands of PCs sold so the numbers didn’t change too much anyway. I still took the adjusted numbers just to secure myself against nitpickers 😉
Apple:
Q1/2008 http://www.apple.com/pr/library/2008/04/23Apple-Reports-Record-Second-Quarter-Results.html
Q2/2008 http://www.apple.com/pr/library/2008/07/21Apple-Reports-Record-Third-Quarter-Results.html
Q3/2008 http://www.apple.com/pr/library/2008/10/21Apple-Reports-Fourth-Quarter-Results.html
Q4/2008 http://www.apple.com/pr/library/2009/01/21Apple-Reports-First-Quarter-Results.html
Q1/2009 http://www.apple.com/pr/library/2009/04/22Apple-Reports-Second-Quarter-Results.html
Q2/2009 http://www.apple.com/pr/library/2009/07/21Apple-Reports-Third-Quarter-Results.html
Q3/2009 http://www.apple.com/pr/library/2009/10/19Apple-Reports-Fourth-Quarter-Results.html
Q4/2009 http://www.apple.com/pr/library/2010/01/25Apple-Reports-First-Quarter-Results.html
Q1/2010 http://www.apple.com/pr/library/2010/04/20Apple-Reports-Second-Quarter-Results.html
Q2/2010 http://www.apple.com/pr/library/2010/07/20Apple-Reports-Third-Quarter-Results.html
Q3/2010 http://www.apple.com/pr/library/2010/10/18Apple-Reports-Fourth-Quarter-Results.html
Q4/2010 http://www.apple.com/pr/library/2011/01/18Apple-Reports-First-Quarter-Results.html
Q1/2011 http://www.apple.com/pr/library/2011/04/20Apple-Reports-Second-Quarter-Results.html
Q2/2011 http://www.apple.com/pr/library/2011/07/19Apple-Reports-Third-Quarter-Results.html
Q3/2011 http://www.apple.com/pr/library/2011/10/18Apple-Reports-Fourth-Quarter-Results.html
Yeah, nothing for the computer manufacturers to worry about in your new graph.
That wasn’t my point. Apple’s Q4/2011 sales for the iPad are projected at 15 million at least.
They WILL have the top market share – and my graph reflects that.
I just found the Deutsche Bank graph to be overly sensational. If you count the iPad as a computer you have to start counting it the day it was first sold not just now and an Analyst should know that.
Are you kidding? This is written by an analyst for a bank, not tech enthusiasts. I’d be concerned ( not worried) if I was Acer or Dell hovering around the 10% range with an average selling price of less than $500, where as the the average selling price of Macs are about $1300 and the average selling price of Macs plus iPads is around $850.
Apple at 17% including iPads, means they are selling 70% more devices than Acer or Dell. All of those devices ASP is at least 50% more than of Acer and Dell ASP.
I think the analysts entire point was to show that even though Apple has a small market share of the PC industry, if you include iPads which sell for a minimum of $500, Apple is making a boatload of money. The original graph was made to emphasize that point, I don’t think that there was any intention to mislead. If the analyst would have just graphed 3Q11 without and with iPads, no one would be complaining, except tech enthusiasts arguing the definition of a PC. But then there would not be much of a graph, for people who like graphs like investors and wall street analysts 😉
This doesn’t even get into the details that most PC vendors need to sell between 4-5 PCs to make the same amount of profit that Apple makes from selling one Mac. An iPad probably has a profit margin of selling 2-3 PCs.
Apple, like them or hate them, are making a lot of money. I wrote is post to counter Brian who seems to think that there is no market disruption.
The principal error the analyst made was to repeat the Q3 column of the graph; where he added the iPad numbers on top of the Mac numbers.
Instead he should have simply introduced one additional data series (i.e. line), not an additional column. Even if the additional data series didn’t start as far back as the others (which would have taken more time than he apparently had).
The additional data series – even foreshortened – would have intuitively shown the relative size/impact of the combined Mac+iPad numbers without crapping all over the rest of the quantitative facts represented by the rest of the graph.
Nice work!
I think it would have been a lot less egregious if he’d simply made his point
I think the analyst just made a single error that explains all of your issues (the „drama“ of Apple’s lift in the last column, and the steepness of its competitors falloff in that same column).
For some reason, the analyst created a graph that was good/normal up to the last column. But for some reason
Sebation,
While the DB graph might be wrong, its also has a very different axis then yours. The only data point that includes the iPad is the last one. All the others ones are just Mac sales which explains why they don’t include it. Also he lists Q3 twice, first without and then with the iPad. No idea why he did that, probably laziness as you said but it might not be as wrong as you think. Obviously you can’t have an x-axis that is supposed to measure time and then repeat a quarter. Just wanted to point that out.
„While the DB graph might be wrong, its also has a very different axis then yours. The only data point that includes the iPad is the last one. All the others ones are just Mac sales which explains why they don’t include it.“
Which is misleading and wrong. You don’t just change the numerator and denominator for one series and one point in time if you want to produce an appropriate comparison of like trends. That’s the point.
In any of these charts is „count the iPad as a computer“ just shorthand for „count tablets as a computer“..?
I know other manufacturers have tiny sales of tablets compared to the iPad, but if we’re going to add iPad sales to Apple’s numbers, it’s only fair to add the tablet sales to all the manufacturers and I haven’t seen it explicitly stated that anyone is doing that in these analyses.
One interesting aspect of that graph is the fact that Lenovo is on a real tear right now. They’ve doubled their market share in the four years covered by that graph. I assume that the growth is in China as I don’t see nearly as many ThinkPads as I used to. I wonder what their profit share looks like…
The chart is not erroneous per se but it is set up to invite misunderstanding. I certainly was astonished on first viewing. Perhaps it originally appeared in a PowerPoint accompanied with commentary— that’s the problem with free-riding on material from others.
And to those who worry about other tablets, don’t.
I understand what you mean but honestly you can’t just count the iPad as a PC in the last quarter you look at. The iPad isn’t new on the scene.
If I were working for Deutsche Bank I knew that my work was under scrutiny. His job are numbers and I don’t hate him for maybe doing this quick and dirty.
I just hope he’s not angry that I did it long and exact 😉
Still wrong.. You’re comparing shipped PC figures from gatner Vs sold number from Apple.
My graph is a comparison to the Deutsche Bank Graph and I highly doubt that the IDG numbers were computers old and that they got anything other than what I got from Apple.
So the comparison stands. Although I don’t mind arguing about what „shipped“ means (in my book you should use sales which is shipped minus returns and then you have to question Apple if they can already say how many returns they will get three weeks after the sales quarter ends)
For those of us who grew up using slide rules, and thus gained some feel for proper values (for the placement of decimal points), the Fortune graph screamed „WRONG“. My mental reaction was something like: „The real story is a bit like that but it can’t be that drastic.“
Thanks for doing the work of sorting it out.
Since you have the numbers already at your disposal, it’d be interesting to see the a third graph which included the iPad sales and Mac sales as two separate items, if you wouldn’t mind.
Great idea! Once the Q4/2011 numbers are published next week I’ll do a follow up and that is a nice option for a third graph. Thanks.
At the moment I have to wait and see how the site holds up since @daringfireball linked here, hope you understand.
When you do write up that post in the coming weeks, can you also create a second graph with absolute numbers rather than just percentages of marketshare? Oh and a table with precise numbers would be nice also. 🙂 I hope I’m not asking too much, I’m guessing that if you have the data to generate the above graph, then you also have the data to available in a table format too, along with absolute numbers. They would help further clarify the situation.
Sure I have the Excel file here it’s not a problem.
The post was getting long and I wanted the reader to be able to digest this quickly. I also have this need for data that’s why I did this in the first place (and just added the short paragraph about the underlying 90 to 100+ million overall sales thing for people to do a little bit of quick math in their head themselves).
This is kind of the middle ground between just linking to the image as many blogs did and inundating the reader with columns after columns of numbers 🙂
Thank you for your work in correcting them. That’s why banks get in trouble. Not enough watchdogs like you.
German pride mixed with Math major I guess 🙂
Thanks for the compliment, in the end it was just a lot of copying and pasting but I thought someone had to do it.
Great work!
P.S. Your site hasn’t a web clip for iPhone’s homescreen. Hope you’ll do it 🙂
Thank you for the work here — analysts, wtf. Do you know if anyone has done this analysis including smart phones — ie: all consumer computing devices
I collected the data for iPhones and iPods when I searched for the iPad numbers so I got those but currently all this is about „PCs“ and not „Smartphones“.
The criticism that hasn’t come up yet is „You left out Android Tablets“ and you know me I can’t wait for the first commenter who’s going to mention that 😉
(I have an answer ready so lets just wait 😉 )
Would like to see it with smartphones — these devices are becoming variants on a spectrum, thinking about PC’s in isolation is as old world as thinking about tablets or phones in isolation
I get your drivt but you have to draw the line somewhere.
I doubt you would say that people buy Smartphones to replace their Desktop computer or buy a Smartphone instead of a new computer.
I wire money, write blog posts and read the internet on my iPhone too but to be honest it’s way more tedious than on my iPad. Try researching for a blog post on a Smartphone. Of course in the end I can pair my Bluetooth keyboard with both iOS devices but there’s a reason why I have a 24″ monitor and the preferred tablet form factor is 10″ – it’s the sweet spot between too heavy to hold in both hands and maximum screen size.
You left out Android Tablets!
🙂
(I want to see your answer…) 🙂
Easy. Just send me URLs to reliable sources stating shipments for each quarter from Q2/2010 to Q3/2011 and I’ll gladly incorporate them into my next update.
I won’t mind adding another graph where all the Android tablets play the lottery how to color the x-Axis (there that was the joke I had ready) 😀
That’s a good one. Coloring the x-axis. Too bad that there are too many colors needed for all tablets. I could just see colorful snow sitting on the x-axis? (actually a single „all others“ line would still sit on the axis?)
While lots of the Android tablets would compete for the flatline x-axis, there are some companies that at least pushed larger quantities into the market, which counts as „shipped“ and that is what the metric of the graph is (by Gartner’s choice). Therefor we would see some dents in the last two years at the bottom when HP published the Touchpad, RIM published the Playbook and of course all the Samsung tablets were shipped.
Of course this doesn’t count as „sold“ but again, this is not the metric because you just can’t say how many PCs will be returned to the manufacturer.
Apple is in the great position of being able to even sell faulty equipment back into the market as „refurbished“. None of their iPads likely have significant return rates – you’d best ask for them to publish total losses – which you won’t get out of them. Those are company secrets for every single manufacturer and none of them is going to publish failure rates or loss-on-production, even though we can always guess how much a product costs by deviding possible writeoffs (like the one RIM did) by number of items shipped.
It’s all a wee bit complex and that’s why I didn’t add anything to the graph. There’s a need to stay basic here and with the iPad making up 97% of the tablet market it’s easily justifiable to disregard other pad producers in THIS graph. You see the total amount with the iPad is 103 million in Q3/2011, with all the other tablets it most likely is 103.5 or maybe 104. Even if we add the other tablets to the graph you wouldn’t be able to notice the dents at the bottom due to line thickness because all the other manufacturers produce so much more units.