iPads, price and self-selection

Apple's iPad event was pretty unsurprising. It was obvious that the large one would be speed-bumped and get lighter. It was also obvious that the Mini would get retina at some point - the  only question was whether the supply chain could deliver enough panels now (with some well-informed people suggesting it could not), and the late-November ship date and $400 price point to how close it was. 

The lack of fingerprint scanning in both the new models was a surprise - it may again be a supply problem, but it means that any platform play Apple has in mind (ie payment) will have a smaller install base to launch onto. 

However, the big puzzle is the price the now old Mini is discounted to: $300. This compares poorly to a new Nexus 7, with comparable resolution to the retina Mini, at $230. The Nexus 7 is of course being sold at very low margin by Google, but does the old Mini really need to be $300 rather than, say, $275 or $250? What is Apple thinking?

On a simplistic level, Apple's tablet market share is clearly shrinking. The chart below, taken from a presentation I've been giving in the Bay Area this week, shows my estimates of tablet unit sales. (They're in increasing order of certainty as you go from top to bottom. Kindle Fire is not included, since I lack the grounds to do quarterly estimates, but obviously it is not relevant outside the USA and a few other markets).

Screen Shot 2013-10-23 at 10.42.10 pm.png

Apple's recent sales decline is largely a product cycle issue, but clearly, sales of other tablets are growing fast. And yet Apple is allowing the price window underneath the iPad to become meaningful. Why?

To me, the really interesting thing about this chart is how small Nexus sales are. Here is a good device with a nice clean Android install, sold at a very aggressive price. On Google's own numbers, almost no-one buys it. Why? Why do sales of the Nexus 10 appear to be under 1m units?

Meanwhile, every single data set shows iPad with at least three quarters of tablet use, be it app installs, web use or any other third-party engagement metric you want. Where are all those other, non-Nexus Android tablets? What's being done with them?

What seems to be happening is that if you want the post-PC vision that Apple and Steve Jobs created, you probably buy an iPad, and Apple has a large majority of that market, and hence of the use of devices for that purpose. This isn’t very surprising: the Android tablet app offer remains far behind the iPad in a way that the Android phone app offer does not. 

But there's also another proposition, a $75-$150 black generic Chinese Android tablet, half the price of a Nexus 7. That proposition is also selling in huge numbers, but it appears to come with a very different type of use. 

Why are people buying these? What are they being used for? They're mostly in China (that’s the pink bar above) and emerging markets and in lower income groups in the west. And it seems that they're being used for a little bit of web, and a  bit of free gaming. Perhaps some book reading. And a LOT of video consumption. In fact, one might argue that for many buyers, these compete with TVs, not iPads, Nexuses and Tabs. But regardless of what they’re being used for, they’re not being used the way iPads are used. In effect, they are the featurephones of tablets. 

If this theory is correct, it suggests that Apple's $300 Mini really isn't a competitive problem, because the iPad doesn't yet face a strong competitive threat (quite unlike the iPhone). Rather, there are actually two quite different markets: the post-PC vision, where Apple is dominant, and a ultra-low margin product that’s also called a tablet but which is really a totally different product. 

Android tablet numbers

As part of Google's event last night, it showed the following chart for Android tablet activations. The source is marked as 'IDC, Gartner and internal Google data' - which is odd, because you would think that Google's own data would be enough for this.  

Screen Shot 2013-07-24 at 17.12.34.png

If I eyeball this chart and back out the numbers, the result looks something like this. 

Screen Shot 2013-07-25 at 11.22.41.png

Clearly, there's an interpretation of this that says Android is eating iPad's sales. 

However, there are two problems with this. The first is the cyclicality issue I explored in a post yesterday (just before this chart was used).

The second, and bigger one, is usage data like this, from Chitika. I'm not, as a general rule, a huge fan of ad network traffic data, but this chart shows numbers that pretty much every web publisher also sees - iPad is 80-90% of tablet web use, and (in the USA) Fire is another 5%. Fire isn't included in Android activations. So Android is maybe 10-15% of use. 

 Yet Google claims there are 70m Android tablets. Apple has sold 155m iPads and probably only 125m of those are still in use. So those 70m Android tablets are 35% or so of the install base - but 10-15% of the traffic. 

There are three possible interpretations of this: 

  • These tablets are being bought in emerging markets (but not China, since Chinese devices generally aren't activated and so won't be in these numbers) and not using western sites
  • They're being bought in developed markets and being used much less, or not at all
  • They're being bought and not used for the internet - they're cheap kids' tablets, baby monitors, points of sale devices...  

The truth is probably a mix of the three. But this doesn't easily lend itself to a simple 'people are buying cheap Androids instead of iPads' narrative. In particular, the tiny share of Nexus sales suggests that people buying these devices are much less interested in a good experience than in the lowest POSSIBLE price. 

Apple cyclicality

Apple's results yesterday were actually pretty boring. But one of the more interesting things is the way that the iPad is now starting to become just as cyclical as the iPhone.  

You can see this at a high level in this chart, which shows unit sales for Apple products since the iPhone launched back in 2007.  

Screen Shot 2013-07-24 at 12.03.18.png

There are a bunch of dynamics here.

  • The iPod, obviously, spikes each Christmas as a gift, though that is declining as a percentage of sales as the overall business declines. 
  • The iPhone has spiked, to an increasing extent, with new product launches
  • The iPad has been driven both by new product launches and by buying seasons
  • Apple changed the iPad product cycle in 2012, from launching in the Spring to launching in the December quarter
  • And, just to complicate things, Apple has shifted from launching new products in China in the March quarter (i.e. after the USA and in the New Year season) to launching them in the December quarter with everything else.

This product cycle impact is clear in the two charts below, which break out iPhone and iPad sales by the primary product that was on sale (older models continue on sale but we don't have a breakdown). 

Screen Shot 2013-07-24 at 12.08.06.png
Screen Shot 2013-07-24 at 12.08.40.png

The iPhone dynamic is pretty obvious, but the iPad cycle is more complex, because sales are being driven by both the calendar season and the age of the product.

For the first two cycles Apple launched the new iPad in the June quarter and sales then rose in the September quarter (back to school?) and again in the Christmas quarter (presents) before falling in March (old product, new one coming). But then in 2012, iPad Mini rumours led to sales falling in September and then Apple launched the Mini and a 4th gen model in the December quarter. So sales shot up in the December quarter, held up well in March (newer product plus Chinese new year) and then slipped this quarter as it's now a nine month old product.

I should point out, incidentally, that competition is a relevant but not sufficient factor to explain these swings. Certainly, Android tablets are having a minimal impact in the West. The iPhone picture is more complex, but the chart below makes it pretty clear that substitution from iPhone to Android cannot be the main factor. 

Screen Shot 2013-07-24 at 16.52.21.png

All of this is a long-winded way of saying that sales patterns for the iPad have yet to settle down. It looks like September quarter iPad sales will be down again, but really, who knows? The one thing that's certain is that there'll be another big spike in December. And that historic comparisons just don't work. 

The broader issue for investors is that the annual cycle means you can't really tell how Apple is doing from the mid-cycle sales numbers. So much of the revenue and the momentum is determined by the December quarter that you really have to wait. 

There is, of course, a significant set of financial implications for Apple from such volatility in production and sales, but that's another topic. 

UPDATE Google released tablet numbers for the first time yesterday: see a discussion here.  

Wired and the iPad

Wired is the magazine that should be selling more than anyone else on the iPad. And it has executed… fairly well, up to a point. So how is it selling?

Well, according to Conde Nast, after the initial surge of curiosity pushed sales over 100k, monthly sales have settled at around 32,000. At $3.99 an issue and allowing for Apple’s 30% commission, that works out at just under $1.1m a year, plus advertising. So, they’re probably covering their production costs, but not a huge amount more. Doesn’t sound too inspiring, does it?

But what broader conclusion should we draw from this? That magazines don’t work on the iPad? Not at all. Take a look at this chart (all the numbers are from Conde Nast and pretty easy to find). 

(The US is a subscription market for magazines, which has caught out some European analysts.) 

85% of Wired readers subscribe. Subscribers can’t use their subscription to get access to the iPad edition - they’d have to pay twice. And since there’s little or nothing extra in the iPad edition that you can’t get in the print title and the website, they don’t. In other words, almost everyone who cares about Wired already subscribes, and most of those people won’t buy the iPad app. And so, the fact that Wired is only selling 32k copies on the iPad right now tells us precisely nothing about demand for magazines on the iPad.

Indeed, the fact that iPad sales are 40% of newsstand sales suggests pretty high adoption amongst impulse purchasers.

Coming soon: an agreement between publishers and Apple that may address this. But tangentially, I suspect that today, Conde Nast may not even have a system that could give print subscribers a free login to an iPad app, even if Apple let them.