Thoughts on Samsung’s Q3 2014 earnings

Note for new readers: this post is part of a series on major tech companies’ earnings for calendar Q3 2014

I’m going to link here quickly to two of my past posts on Samsung as they provide useful context which I don’t want to revisit in this post:

  • Is Samsung’s Exceptionalism Coming to an End? – on Techpinions. This post focused on the fact that Samsung had always seemed an exception to my rules for success in tech, and it seemed inevitable that its reign would sometime come to an end.
  • Thoughts on Samsung’s Q2 2014 earnings (and its future) – on this blog. In this post I talked about Samsung’s margins as being unusual for a pure consumer electronics business, and how those margins were likely to revert to the mean eventually because of Samsung’s lack of differentiation. (This is also the most-read of any post on this site since its inception a year ago)

All past analysis on Samsung on this blog can be seen here. My analysis tends to focus on Samsung’s mobile business specifically – other than a brief mention of chips, I won’t cover Samsung Electronics’ other businesses (air conditioning, TVs, refrigerators and display panels, among other things).

Plunging growth numbers

It’s worth starting with a quick financial review, to make sure we’re all on the same page. The relevant business unit from a mobile point of view is “IM”, or IT and Mobile communications. This includes PCs and tablets as well as smartphones, but it’s the lowest-level division where Samsung reports both revenues and profits, so it’s the one we’ll focus on here. However, it does report revenues for the Mobile segment specifically, so we’ll start there. First, revenue growth:

Screenshot 2014-10-30 13.51.42

As you can see, year on year revenue growth has plummeted from positive 90% in Q1 2012 to negative 30% this past quarter. The decline over the past year is alarmingly linear, and one wonders how much worse things can get before that decline levels off. Continue reading

Thoughts on Twitter earnings for Q3 2014

Note to new readers: this is part of a series on major tech companies’ Q3 2014 earnings. I do similar analysis each quarter for certain companies. Previous pieces on Twitter may be found here

Last quarter I talked a little about the challenge of Twitter’s user growth, and what I uncovered as its renewed emphasis on logged-out users (a theme Dick Costolo had first talked about three years ago and to which he has now returned). I and others have talked about Twitter’s different user groups and on yesterday’s earnings call Costolo explicitly spoke about three concentric circles (he actually said eccentric, but I’m fairly certain he meant concentric) representing Twitter’s users. That’s something I’ll return to below. There will be three major points to this analysis:

  • Twitter has given us metrics to measure its growth and potential, but it keeps backing away from or hedging on these without providing better ones
  • Twitter’s core user growth has slowed in recent quarters, and though it wants us to think about a broader group of users, this broader group is both ill-defined and much harder to monetize
  • Twitter is very successfully managing the same transition to mobile advertising as Facebook, but it’s not yet clear how sustainable this business will be for either company.

Twitter’s metrics aren’t telling a good story

The three main metrics Twitter has established for tracking its progress are:

  • User growth: as measured by growth in monthly active users (MAUs)
  • Engagement: as measured by timeline views per MAU
  • Monetization: as measured by ad revenue per thousand timeline views.

These three may be seen as levers, or multipliers, for growth. To the extent that all three are improving, they should drive faster and faster growth for the company, as each change in the first is magnified or multiplied by growth in the second, which in turn is magnified or multiplied by the third. The challenge is that of these three metrics, only one is heading in the right direction, and that’s ad revenue per thousand timeline views:

Ad revenue per 1000 timeline views

As you can see, international monetization continues to lag significantly behind US monetization, which is a function both of Twitter’s underdeveloped overseas sales channels for ads and the lower ad spend in every other market around the world. But both sets of numbers are moving in the right direction, and in the case of US revenues very rapidly so. I’ll come back to the drivers for this growth later, but this is the single best sign about Twitter’s business: it’s able to monetize what usage there is better and better, and this in turn has driven strong overall growth in ad revenues (in thousands of dollars):

Twitter ad revenue

The challenge is that this growth is strongly tied to direct sales, which is an expensive proposition in its own right. Even as some of Twitter’s other cost lines have shrunk as a percentage of revenue, sales and marketing costs are rising over timeTwitter S and M costs as percent of revenue

Note that this line excludes stock-based compensation, which adds significantly to the overall cost of these efforts. With SBC included, sales and marketing costs are equivalent to 45% of revenue. Now, Twitter does have self-service channels for selling ads, but they’re not filling up to the extent that they are making a meaningful dent in the overall ad load – the vast majority of Twitter’s ad sales are still made directly. That has to change to make this growth more sustainable and less expensive, which is critical to Twitter’s long term growth and profitability. It is rolling out those self-service tools to more markets, but it isn’t making much of a difference yet.

User growth – the foundation of overall growth – has slowed significantly. The chart below shows sequential growth in MAUs over the past few years, and it added just 13 million this quarter:

Sequential MAU growth

As you can see, with the exception of Q4 last year, which was a monumentally poor quarter for user growth, Twitter hasn’t added fewer than 13 million in a quarter since the fourth quarter of 2010. Growth has slowed significantly in recent quarters from its peaks in 2011 and 2012, and this is a major worry. To put things in context, Facebook has well over a billion MAUs and growing, while Amazon – obviously a very different business – has roughly the same number of active customer accounts as Twitter has MAUs. My point here is this: if you want to be a big business, you can either be a billion-plus-user company like Facebook and monetize each user modestly, or a quarter-billion-user company like Amazon and monetize them much more aggressively. But you can’t be a quarter-billion-user company and monetize them modestly and still be a big company. Hence the focus on other user metrics which I’ll come back to later.

As far as engagement goes, the story’s unfortunately bad there too, especially in Twitter’s most important market, the US:

Timeline views per MAU

A few quarters ago, Twitter explained that it would see some softness in this number in the short term because of changes to its apps, which would cause the number of timeline views to go down even as engagement improved. However, we should be long past those effects and yet timeline views per MAU are still falling in the US quarter-on-quarter and year-on-year. International timeline views per MAU have flattened too in the past quarter, another worrying sign.

New metrics needed, but Twitter won’t provide them

I’ve focused so far on the metrics Twitter has in the past suggested we should use to measure their progress. However, on both last quarter’s earnings call and yesterday’s call, the company recited a variety of other data points which it hopes will tell its story better. But it isn’t yet providing this data consistently, and it isn’t convincingly explaining why the existing metrics look so bad.

One good example of this is Costolo’s renewed emphasis on logged-out users. I decided to make his concentric circles concept into a concrete diagram, and add some detail which I think is critical to understanding the limitations of this argument:

Twitter concentric circles

The point Costolo makes is that the true Twitter audience is much broader than the MAU number it’s focused on for so long, because of the two outside circles. The key challenge, though, is that although this audience is much bigger, it’s much harder to monetize, because Twitter knows far less about these users and can track them far less effectively, which makes them much less attractive to advertisers. Given Twitter’s challenges in filling its ad inventory even for its MAUs, how is it ever going to increase load for these other groups? I’ve talked before about the most effective forms of online advertising, and the challenge for Twitter is that much of these larger user groups falls into the less attractive categories below:

Advertising relevance vs timeliness

Twitter’s mobile growth is impressive, but is it sustainable?

Twitter is beginning to follow a similar trajectory to Facebook in its impressive mobile growth. Of course, Twitter has always had a significant mobile component, but its transition in this direction is still impressive. Twitter doesn’t directly report mobile advertising and non-mobile advertising, but it provides enough information that we can derive them with some accuracy, as shown in the chart below (the numbers are in thousands of dollars):

Twitter mobile vs non mobile ad revs

As you can see, mobile advertising is ramping strongly, while non-mobile ad revenues are flat or declining. This is a very similar picture to Facebook’s, and the components of this growth are starting to be similar too. Twitter’s executives specifically called out the contribution of mobile app install ads as a driver on yesterday’s earnings. I’ve talked before about the questionable sustainability of app install ads as a growth driver, and I continue to worry about that as a driver for ad revenue for Twitter and other companies investing aggressively in this space. But more importantly Twitter’s syndicated and logged-out usage is likely to be heavily browser-driven, rather than app-driven, and there will be far less scope for advertising in those environments than in the app environment. The mobile ad revenue growth is fantastic, but Twitter’s wider user groups are far less likely to engage in the Twitter app, which means Twitter needs ad products that will work better on both mobile and desktop browsers.

Amazon and the cost of growth

Note to new readers: This is part of a series on major tech companies’ Q3 2014 earnings. Prior analysis on Amazon can be seen here.

Last quarter I did a pretty deep dive on Amazon and many of its financial metrics, and specifically addressed the question of how sustainable its business is, and whether it’s really realistic to expect it to be able to turn on the “profit spigot” at some point in the future. This quarter I’m taking a simpler, more focused approach on the cost of growth, the pursuit of which is the core reason Amazon is losing money.

Hiring at a fierce rate

Amazon’s hiring new employees at a fierce rate. This is one of the two primary costs of having a very logistics-heavy business, and the key driver behind Amazon’s investments in robotics. But for now the company still needs lots and lots of people to staff fulfillment centers around the world. Here’s are several charts showing statistics related to the number of employees:

Number of Employees

As you can see, the growth is enormously rapid. I pointed out yesterday in a tweet that Amazon hired more than twice as many employees just in the past quarter than Facebook has employees overall. They’re obviously very different businesses, but this just highlights quite how people- and infrastructure-intensive Amazon’s business is. Amazon has now reached a rate of 40,000 net new employees per year, which is astonishing when you consider it only employed 40,000 in total back in 2011: Year on year employee growth

The other thing about Amazon which is different from many of its major competitors is that the kind of people who make up the employee ranks are very different. At Amazon they’re generally far less skilled and therefore much cheaper to employ, which means that it can afford to drive far less revenue per employee than companies such as Apple, Microsoft and Google, which employ a much more expensive mix of employees: Continue reading

Thoughts on Microsoft’s Q3 2014 earnings

Note to new readers: I use calendar rather than fiscal quarters for easier comparability between companies’ results. As such, I use Q3 2014 universally to refer to the period ending September 2014, even though some companies (in this case Microsoft) use a different reporting calendar). All the charts and analysis below use calendar rather than Microsoft fiscal quarters. This is part of a series on major tech companies’ Q3 2014 earnings. Prior analysis on Microsoft can be seen here

There are four sections in this post – click below to jump to each of them:

Consumer hardware performing surprisingly well

One of the most surprising things about today’s earnings was that the consumer hardware businesses, which are important to Microsoft’s strategy but have performed poorly, all got a bit of a boost this quarter. Xbox, Surface and Lumia devices all had strong year on year growth:

Lumia device sales Surface revenue Xbox unit salesAll three device categories saw record results, in fact – Lumia sales and Surface revenue were both the highest they’ve ever been, while Xbox unit sales were the highest they’ve ever been outside the seasonally much higher fourth calendar quarter. Now, we have to put all this in context: both Lumia and Surface sales are dwarfed by the respective markets in which they compete. But on a relative basis, the company continues to see growth in both categories, and the Xbox year on year growth was solid too.

Now, much has been made of the fact that Microsoft reported a positive gross margin for the first time for the Surface as well. Long-time readers will now that I have teased estimated gross margin numbers out of Microsoft’s SEC filings for the past few quarters, so I thought I’d provide an update on the same basis. Here’s the chart including this quarter’s numbers:

Surface financials

As you can see, the number does indeed seem to be positive this quarter on the basis of my analysis, at a little under $100 million. That’s a gross margin of around 9%, which is not earth-shattering and in fact about half the gross margin of the phone business at Microsoft. But it’s progress, as with the growth numbers above. There’s a long way to go to get to the kind of gross margins that would lead to true profitability once marketing and other costs are factored in. How many Surface devices did Microsoft sell in the quarter? Well, they won’t say, but given the new version starts at $800, it’s entirely possible that the company sold a million or fewer Surface tablets in total, and likely well under a million Surface Pro 3s in their first full quarter on sale. It sold about ten times as many Lumia devices, and about 40-50 times as many mobile phones, just for comparison’s sake.

Cloud run-rates and actuals

Much was made last quarter of Microsoft’s cloud “run-rate” which turned out to be based on a single month of sales, somewhat unconventionally. Here, too, I’ve estimated cloud revenues in the past and will do so again here. Here are my estimated numbers for Cloud and Enterprise Services, the two major components of the Commercial Other segment at Microsoft:

Cloud run rate

Last quarter, the run-rate was said to be $4 billion, and I pointed out that it wasn’t on the basis of the full quarter’s results, but that the trajectory was clearly there to support that claim. This quarter, the number did indeed cross $1 billion (it was just under $1.2 billion), so the run-rate now on a quarterly basis is almost enough to carry it to $5 billion for the year. This business clearly is growing very quickly, and we’re getting to some very big numbers. Last quarter it looked like Cloud Services might overtake the older and slower-growing Enterprise Services category for the first time. It didn’t – mostly because Enterprise Services grew faster – but almost certainly will next quarter.

Office in transition

Microsoft used to report a year on year growth rate for Consumer Office, but it was based solely on legacy Office sales to consumers, and excluded the newer Office 365 versions for consumers. This quarter, it changed its reporting to reflect the overall picture, which is actually fairly reassuring. The combined number is growing ever so slightly year on year, as shown in the chart below (again, the numbers are my extrapolations from Microsoft’s SEC filings):

Office in transition

This is a somewhat cyclical business, so the key is to compare Q3 with last year’s Q3 rather than looking at quarter-to-quarter changes. Office 365, of course, is not cyclical in nature, since it’s subscription-based rather than a one-off sale. But it’s still much smaller, so it doesn’t affect the overall line much.

One interesting thing to note is that Office 365 Home and Personal subscribers grew strongly, from just 2 million a year ago and 5.6 million last quarter to 7.1 million this quarter. However, revenue from those subscribers seemed to barely budge as far as I can tell (I have a couple of different ways of arriving at this figure and they end up in almost exactly the same place). The number seems to have stayed at around $125m for the quarter from last quarter to this, despite the growth in subscribers. There are two possible explanations, which I think both contributed to this: Microsoft has begun a strategy of giving away some of its paid services with new hardware purchases, so some of these subscribers may not be paying subscribers (at least not yet). In addition, the Personal version of Office 365, which retails for $70/year instead of $100/year, went on sale shortly before the quarter began, so that may be starting to drag down the average price users are paying. It’s also worth remembering that these sales are the best indicator of how much people are using the full versions of Office for iPad. The conclusion: not that many.

Online advertising divergence continues

As with Google and Yahoo, Microsoft’s online advertising business continues to see divergent trends around search and display advertising. Here are my estimates for these two businesses:

MS online advertising divergence

As you can see, Display advertising revenue, which was almost as high as Search revenue two years ago, has fallen to a fraction of it, while Search revenue continues to grow reasonably strongly year on year. Bing and the Yahoo arrangement seem to be working well for Microsoft, but it’s struggling to make a dent in Display advertising despite revamped MSN, and other products over the past year. Given Microsoft’s strategy of attacking Google over its use of advertising to fund its products, I’m starting to wonder whether Microsoft shouldn’t just strip out the ads from some of these products and run them at a loss as platform differentiators, much as Apple bundles many of its services into purchases of devices.

That’s it for now – I may have more later in this busy earnings period. Amazon should be next up, likely tomorrow.


Thoughts on Apple Q3 2014 earnings

Note for new readers: I stick to calendar quarters in analyzing earnings, because it makes cross-company comparisons easier. As such, all references to “Q3 2014″ refer to the quarter ending September 2014, not Apple’s fiscal Q3 2014. Once again, this post is part of a series on major tech companies’ earnings (this is the second for the current round of earnings; analysis of Q1 2014 earnings can be found here, and for Q2 2014 can be found here). 

iPad plus Mac is the number to watch

Lots of attention was paid by both Apple and commentators this quarter to both the iPad and Mac numbers, but the key is increasingly to look at the combined results of these two segments, both in unit shipment and revenue terms. For a while there, as the Mac started to shrink, iPad growth offset that decline, but the two have now switched places. Nonetheless, year on year, the two combined continue to operate within a very narrow range:

Mac plus iPad revenues Mac plus iPad shipments

ASPs continue to tell an interesting story

Given the near demise of the iPod, and the declining average selling price (ASP) of the iPad, the iPhone and Mac are actually the best two products for Apple to be selling. Interestingly, they’re also the two products which are most likely to experience substitutional effects with the iPad. Note the ASP trends in the chart below, which shows trailing 4-quarter ASPs to smooth the quarter-to-quarter fluctuations:

Apple ASPs

The iPhone ASP continues to hold up remarkably well, staying within $10 of $600 for the last five quarters, though rather lower than the $640 or so it regularly hit earlier. With the launch of the iPhone 6 Plus it’s likely that ASPs will trend somewhat higher over the next couple of quarters, and this quarter’s ASP was $42 higher than last quarter’s, partly for that reason.

Who’s buying iPads?

I’ve done some analysis previously on iPad replacement cycles and theorized that we’re due for faster sales over the next few quarters as the many iPads sold 2-4 years ago become due for upgrades. I thought I’d revisit some of those numbers in the context of what we heard today to answer the question of who’s buying iPads. Tim Cook has now given several data points as to the percentage of iPad buyers who were first time buyers, and it’s interesting to look at what that data signifies. Using that data, I’ve put together estimates of the breakdown between new buyers and upgrade buyers of iPads, as shown below:

iPad Buyers By Source

If these numbers are correct, they appear to show two trends:

  • The number of people buying upgrades to existing iPads is rising slowly in a cyclical pattern, as would be expected with a growing installed base (I’ll come back to this below)
  • The number of people buying new iPads has been extremely cyclical too, but appears to have slowed somewhat over recent quarters. Though the Q4 peak last year was higher than the one the year before, every quarter since has been off the year-earlier quarter by some margin. Though Tim Cook rightly points out that the market isn’t saturated, the number of new buyers does seem to be falling somewhat.

Let’s look at that upgrade number in the context of the installed base (the caveat here is that we’re working with two sets of estimated data now). My guess is that the percentage of the existing iPad base that upgrades in any given quarter is around 3%, with a higher number in Q4, both because it’s a big buying quarter anyway and because as a result it’s the anniversary of many earlier purchases. I think the rate in those quarters likely spikes to around 4.5%. On that basis, as the base continues to grow by a few million every quarter, if the upgrade rate holds steady, the number of iPads sold to existing owners will continue to grow steadily too. The big question then becomes whether Apple can turn around the new buyers number with the new iPads it launched last week, and the lower prices on older iPads. I suspect it will and we’ll see a really big fourth quarter, perhaps the first in a year that’s higher than the year-ago quarter.

Apple Watch buried among Other Products

One of the more interesting tidbits on the earnings call was not a financial data point in its own right, but an indicator of future reporting changes. Namely, that the Apple Watch will be reported under Other Products along with both the existing Accessories business (including Beats) and the iPod business, which was formerly reported separately but is becoming so small as to be no longer worthwhile reporting in its own right. It’s worth looking at the history of other Apple products here for a precedent:

  • The iPod launched in October 2001, and in the next earnings release Apple reported the number of shipments, but didn’t break the product out in its Data Summary with its Mac shipment and revenue data until two years after launch, in October 2003. The company continued to provide occasional iPod shipment numbers in the interim, however.
  • The iPhone and iPad both received immediate status as segments in their own right immediately after launch, with a full revenue and shipment breakdown (though in both cases muddied by the fact that other related revenue was lumped in with device sales revenue for a time).

What does it signify that Apple won’t report Apple Watch shipments and revenue in full detail from the outset? I think two things:

  • Apple is exhibiting caution ahead of what is in some ways its most unpredictable new product category in many years, since the iPod. Apple as a company has many times more customers today than it did then, but the Apple Watch is as big a departure from its current product line as the iPod was in its time, and it’s inherently difficult to predict how many it will sell. As such, the lack of reporting in the short term may reflect an abundance of caution about breaking out a nascent category.
  • Apple’s leadership alluded to this on the earnings call, but the Apple Watch will also have a far more diverse set of price points than any of its other products, ranging from $349 to several thousand dollars, and as such the average selling price will be a huge clue as to which models are selling in a way that it has never been for the iPhone, iPad or even the Mac. As such, Apple is keeping this commercially sensitive data out of competitors’ hands, at least for the time being.

However, all that said, within two years of the launch of the iPod Apple was providing a detailed breakout of both shipments and revenues, and I’d very much expect that if the Apple Watch sells at all well we’ll get (a) ad-hoc reporting of key metrics such as shipments right from the start as with the iPod and (b) a full breakout at such a time as the Apple Watch becomes significant enough as a revenue generator to warrant its own segment. With an ASP that’s likely to be in the same ballpark as the iPad or higher, it will only have to sell a few million to become material to Apple’s earnings overall, and I would expect that to happen within the first few quarters. It will be hard for Apple to keep these numbers buried out of sight for very long.

I’ll likely do another post or two this week as I continue digging through the numbers, so that’s it for now.


Quick Thoughts: Apple SIM

Yesterday afternoon, amid a flurry of tweets and articles talking about how revolutionary the Apple SIM is, I posted this:

I wanted to take some time to elaborate on that thought and on the Apple SIM in general.

A revolutionary model

First, let’s give credit where it’s due, to both Apple and the three participating US carriers and the UK’s EE: this is a revolutionary model, and all these companies deserve kudos for being willing to innovate and try something different, despite the risks to the carriers in particular. Why might the carriers be willing to go along with this? Well, for one, they risk being left out if they don’t, as Verizon is here in the US (likely because of its overall skepticism about the new business models for device sales). But secondly, all the US carriers would like to see a much higher proportion of tablets sold come with cellular connectivity. Today, the cellular attach rate is very low, and only part of that is down to the $130 premium Apple charges for LTE devices. The rest is down to the fact that buying one of these devices has meant committing to a particular carrier and potentially a long-term contract before you even know how much use you’ll get out of the device.

This new model means you can go into an Apple store and walk out with a cellular-enabled iPad, and start using the cellular function immediately, without making any sort of commitment to a carrier at all. You can try out their service, switch carriers easily, and determine whether or not you want to continue to use the service, all without ever going near a carrier store. The downside for the carriers, of course, is a lack of lockin, though I suspect once an iPad customer decides which carrier to use they may well end up moving the device onto a family or shared data plan for a better deal than they’ll get otherwise.

Beware of assuming the same model works in phones

Having said all that, I think the clamoring over this revolutionary new model is over-done when it comes to smartphones. There are at least two fundamental differences between smartphones and tablets in this context:

  • The vast majority of people buying a smartphone already have one, running on a specific carrier
  • The vast majority of smartphones are bought with financial assistance from a carrier, either the old subsidy model or the newer financing models. As such, almost no-one pays the full cost of an iPhone up front.

The Apple SIM model works so well in the iPad context precisely because most people aren’t getting financial help from their carrier to buy a device, and because many of these customers don’t have a tablet connected to a carrier yet. They’re in experimentation mode, and the Apple SIM model works perfectly as a result.

The problem with applying this model to iPhones is that the carriers serve this other critical function of helping to reduce either the upfront or total cost of the device, and Apple doesn’t (yet) provide an alternative. As such, the model would only be applicable for people who were willing to pay the full cost of the phone up front, which is an entirely marginal market. However, if Apple were to start offering financing or leasing plans for iPhones as I’ve described elsewhere previously, then the Apple SIM model would make a lot more sense. Oh, and by the way, both of the carriers’ major holds on customers (the contract and network lockin) would be broken at once, which would be enormously disruptive.

Apple is entirely capable of pursuing this kind of model itself. This could be either the carrier financing model, with the cost of a phone spread over a 12-24 month period, or an “iPhone for life” program under which a customer pays a fee each month to always have the latest iPhone model. Under the latter model, the older device would be handed back to be refurbed and resold when the customer gets a new phone. Apple has the deep pockets to fund such a model, and it would help to smooth out its revenues across the year too even as most of the upgrades continue to happen in the third and fourth quarters.

In short, the Apple SIM is a step in the direction of a new relationship between Apple customers, Apple and the carriers. But in order to reach its full potential in the iPhone context, Apple needs to make another significant change: allowing customers to spread the cost of owning an iPhone over a longer period. Only if it does that will the Apple SIM be truly disruptive.

Further thoughts on iPad sales

On the morning of Apple’s latest iPad event, I wanted to quickly revisit the topic of iPad sales and share an idea that’s come up again and again as I’ve discussed the topic with other analysts and with reporters in the runup to today’s event. I’ve posted the video below to YouTube as a way of illustrating this idea visually, and the text below is largely the transcript of the video, with some of the same images used to illustrate key points. I’d love your feedback on the video format and on the blog post, as always. Feel free to connect with me on Twitter at @jandawson or to email me at

(you may want to watch the video on to see it bigger, and whether you watch it here or there I suggest switching to the highest available resolution.)

I’ve done two blog posts recently which made use of some version of this diagram:

Apple product evolution

The first looked at the Apple Watch as the latest in a long line of increasingly mobile and personal computers Apple has released since the first Apple computers. The second examined the question of how many computers we actually need, and the tension that exists when we end up purchasing several of them to accomplish the same set of tasks in slightly different ways.

I wanted to return to this diagram to illustrate an idea that’s been percolating in my mind since I first drew this diagram, and that’s the iPad’s place in this evolution.

One way to see this diagram is as the technological equivalent of this hackneyed picture of human evolution:


But of course there’s a problem with that. If you play back the Apple version of this evolution there’s a historical quirk – the iPad didn’t arrive at its logical place in the evolutionary chain: it was late:

iPad late

Though Apple started work on the iPad before the iPhone, it came to shelve that project and focus on the iPhone instead, only returning to the iPad later. As such, in a historical quirk that would have been impossible in a true evolutionary progression, the iPhone’s logical progenitor ended up coming later in the evolutionary chain. Steve Jobs explicitly recognized this relationship in introducing the iPad by placing it between the MacBook and the iPhone in Apple’s product lineup.

The problem with that approach is that it’s created a strange set of expectations for what the iPad should do, both as a device to be used and as a member of Apple’s product portfolio. Had it launched first, with the iPhone launching later, it would have been natural to assume that the iPhone would eventually cannibalize it much as it did the iPod. But because the iPad launched after the iPhone, it created this unnatural expectation that it would complement the iPhone and perhaps even exceed its success.

Only when you see the iPad in its natural place in the evolution of Apple products – despite the actual timing – does it start to become clear where the iPad sits and what its future might hold. The iPhone – and not the iPad – is the culmination of this evolution, with the Apple Watch the next evolutionary step (with the potential eventually to become the pinnacle of this evolutionary process, in time replacing the iPhone).

What does this mean in terms of iPad sales? Well, there are two big questions we don’t know the answer to with iPad sales. The first has been well discussed, and was the topic of one of my earlier blog posts, and that’s replacement cycles. With a product that’s just four years old, and which didn’t start to sell in really large numbers until 2011, it’s very hard to calculate what those might be. But it’s easy to imagine that they’re longer than iPhone replacement cycles, perhaps as long as three or four years. As such, I’ve argued that we might see a major upgrade cycle over the next couple of years as that first big wave of iPad purchases in 2011 ages to the point where it needs to be replaced.

The other big question, though, which hasn’t been discussed nearly as much, is the degree to which some iPads won’t be replaced at all, because their owners stop using them entirely. Because of the historical evolution we’ve discussed, and especially because of the launch of larger and larger smartphones, now including the iPhone 6 Plus from Apple itself, there will be many people who no longer feel the need for an iPad at all, especially the only slightly larger iPad Mini. So the real question splits into three parts: firstly, how many of the iPads Apple has sold are still in use? Secondly, what percentage of those will be upgraded or replaced at all? And thirdly, how quickly will those upgrades happen following the initial purchase?

All of which brings us to Apple’s event today, at which it will launch refreshed iPads. Apple’s job with regard to the iPad during this event is threefold: give the many existing owners a reason to upgrade, give people who haven’t yet tried iPad a reason to buy their first, and give people who may have abandoned an earlier version of the iPad a reason to give it another try. The first of these is in some ways the easiest – the new features (Touch ID, possibly a higher resolution screen) and the requisite spec bumps will increase the gap in performance between the iPads 2 and 3 many people own and the latest device even further, making an upgrade more compelling. And Apple has actually been doing a good job bringing more and more converts to the iPad too, with around 50-70% of buyers being either new to iPads or to tablets as a whole in the last few quarters. It’s the third task that’s the toughest, and I wonder to what extent Apple should even be trying to convince former iPad users to come back. If an iPhone 6 Plus is the right device, and obviates the need for an iPad, should Apple merely embrace the survival of the fittest device here?

This, in my mind, is the most interesting aspect of today’s event, because it will show us how Apple views the iPad in a product portfolio that now includes the forthcoming Apple Watch as well as the much larger iPhones. Again, had the iPad launched before the iPhone as befits its place in the evolutionary chain, we wouldn’t be marveling at the stagnant and even falling sales over the last year or so. It’s only because of the quirk of timing of the original iPhone and iPad launches that we’re even wondering about this at all. I’m looking forward to seeing whether Apple sees it that way too.

iPhone 6 and 6 Plus thoughts

As I’ve mentioned before, I don’t do “reviews” of devices as such, because I think others who focus on those full time do a better job, and I’ll add little value. But I do occasionally post some thoughts on the devices I spend time with, and I thought I’d do that with the iPhone 6 and iPhone 6 Plus which I’ve been using since they came out, thanks to two devices on loan from Apple.

First off, I get review units of lots of devices, including many Android devices, and so I’m already very accustomed to the larger sizes, unlike many regular iPhone users (several of my family members and friends included). I noted on Twitter that John Gruber’s response as detailed in his review was probably much closer to those of many regular iPhone users than those of most reviewers. As such, I was rather looking forward to the larger sizes, since the iPhone 5S I use when not testing something else was coming to feel very small in comparison. However, I’ve never been a fan of the really big devices – those with over 5″ screens, so I was curious to see how I’d respond to the 6 Plus. I first spent a week with the iPhone 6, and then a week with the iPhone 6 Plus, and have been back on the iPhone 6 since then.

iPhone 6 – what the iPhone was meant to be

For me, the iPhone 6 is what the iPhone was always meant to be – it’s the perfect instantiation of iOS on a smartphone. Just the right size, with a really great weight and thinness, which makes it fit wonderfully in the hand. I immediately liked it better than all the other iPhones I’ve tried. I’d been using the first and second generation Moto X devices over the previous few weeks, and had really enjoyed both, though the new version seemed a little large for my taste, and the iPhone 6 does a wonderful job of providing a great screen size without an over-large device, a great in-between experience between the two Moto X devices.

iPhone 6 Plus – great too, just not for me

I forced myself to use the iPhone 6 Plus for a week as well. It is enormously bigger than the 6, and feels so in every way – in your hand, in your pocket, wherever. That makes it fantastic for certain things – I found myself willing to read things I normally would have turned to an iPad for, and playing games optimized for the larger screen was great fun too (I discovered several fun new ones including the Box Trolls movie tie-in game, Beach Buggy Racing and FIFA 15).  Interestingly, my wife, who’s currently using an iPhone 5 and has never liked any of the larger Android phones I’ve shown her, immediately thought this might be a good fit for her. Her reasoning was that she runs her whole life (and our kids’ lives) from her phone most of the time, rarely being in a situation where she can use her laptop, so the bigger screen would make that easier.

As for me, I never did completely get used to the larger sized device, and could never get quite comfortable with it. It is better than most of the Androids I’ve used in this size range, in that it’s narrower, so it’s easier to get your hand around. But it’s not for me. I found taking pictures with it one-handed particularly difficult – I just couldn’t get the device to balance properly when using it in that way, something I don’t have a problem with on the iPhone 6. I was glad to go back to the iPhone 6 after my week with the 6 Plus.

But all this just highlights something that’s never been the case before with Apple’s iPhone line: it’s always been obvious which device you should get (and which I should recommend) with the iPhones before, but it’s not obvious anymore. The iPhone 6 Plus is absolutely right for some people, including apparently my wife, while the iPhone 6 is a better fit for others. Just as the iPad Air and iPad Mini are better fits for different people, and just as has always been the case with MacBooks and so on too.


I’ve been running iOS 8 since it first became available to developers on my iPhone 5S, so the software here wasn’t that new. But by the time it was released on the iPhone 6es, I found it to be relatively bug-free, with only occasional issues, mostly triggered by apps that hadn’t been upgraded rather than the OS itself. I’ve enjoyed some of the enhancements and upgrades, including improvements to Siri, Spotlight search and so on. I haven’t yet made use of some of the Continuity and Handoff features although they are working on some iPads also running iOS 8. Calls and messages come through on those, but I simply don’t find myself using those features at all. I’ve been running Yosemite on a Mac Pro for a while, but it doesn’t have Bluetooth LE and so doesn’t support Handoff, and I’m looking forward to trying Handoff on a MacBook Air that does support BLE when Yosemite ships.


One of the areas where the iPhone has always led is photography, and I’ve found that the improvements here keep the iPhone above and beyond every other device I’ve tested in this department. I haven’t spent inordinate amounts of time deliberately testing the camera but I do take quite a few pictures in the normal course of events. A gallery of photos from the two devices (most of them raw, with a few edited in Snapseed and/or Instagram) can be found on my Flickr page here. I live in Utah, which is a picturesque place (indeed, Apple shot the test footage for the iPhone 6 launch there), so that helps!


One of the first questions almost everyone has asked when they see or hear that I have the iPhone 6 Plus is “does it bend?” This has been true for friends and family members, the teenagers I work with at Church, and a workman who was installing something at our new house this week. “Bendgate” certainly seems to have captured the popular attention and has unfortunately become one of the first things people think about when confronted with the 6 Plus. I doubt it has stopped many people from buying one, but it’s still striking how often it comes up. For my own part, I haven’t seen any sort of bending with the 6 Plus review unit I have. I haven’t tried extremely hard to bend it, but it was in a front jeans pocket for much of the week I tested it, and it simply wasn’t an issue. I suspect it won’t be for all but the unluckiest users either.


In my mind, the iPhone remains the phone to beat. I test lots of devices, and I really enjoy the better Android phones too (I particularly enjoyed the Moto X I tried recently). But the iPhone is my personal device of choice, and the one I always come back to. The one sacrifice lately has been screen size, and one of my other pet peeves (the lack of a swiping keyboard) has also been resolved with iOS 8’s third-party keyboard support. Interestingly, I haven’t used the swiping keyboards much – I’ve found them too error-prone, and found the built-in predictive keyboard to be pretty good. SwiftKey has just updated its app, and it seems better now, so I may try it some more. But overall, the iPhone 6 and 6 Plus just reinforce the iPhone’s place in my mind as the top phone, and especially when it comes to the camera. There really isn’t anything meaningful you can do with an Android phone now that you can’t do on the iPhone 6 or 6 Plus, and that’s really saying something.

Smartphone margins and Samsung

Since I’ve been talking to a few journalists about Samsung’s fairly dismal Q3 earnings forecast, and pointing people back to my previous pieces on Samsung, one thing that’s been coming up a lot is profitability and margins in smartphones. I thought I’d quickly jot down a couple of thoughts on this point in particular, as there seems to be a poor understanding of this topic in general.

The key here is that there are at least two components to profitability in smartphones, though they’re often conflated or one ignored entirely:

  • Gross margin, i.e. the difference between the cost of manufacturing a device (cost of sales or cost of goods sold) and its selling price.
  • Operating margins, i.e. gross margins minus all the shared costs of running the business that sells the smartphones, including all non-manufacturing employees, advertising, general and administrative costs, depreciation and amortization and so on.

It is entirely possible for a company to make a positive gross margin on each phone sold (even a significant one) and yet be hideously loss-making because its advertising, general and other costs are greater than the gross margin on the device. Motorola, for example, claims that it makes money on each device it sells, but that’s a reference to gross margins, and the business unit itself continues to be unprofitable as part of Google, because the other costs are greater than the modest gross margin it makes on its device sales.

This is important in the context of Samsung’s earnings forecast for a couple of specific reasons. Firstly, Samsung’s overall margins have benefited from two key factors:

  • Firstly, it has achieved significant share of the premium smartphone market, and an extremely high share of the premium Android market. Gross margins are much higher at the premium end than at the low end, and this is a major component of Samsung’s overall high margins relative to every competitor except Apple.
  • Secondly, it has massive scale, as easily the largest smartphone vendor in the world over the last couple of years. Scale is important because it spreads shared costs such as advertising, general employee costs and so on over a much greater number of devices. In other words, if Samsung spent a billion dollars on advertising and sold a hundred million phones, advertising costs per device would be $10, but if it sold two hundred million phones, advertising costs per device would be $5. The more devices over which you can spread these shared costs, the bigger the impact on operating margins. There are also some scale benefits to gross margins, of course.

Both this quarter and in previous quarters, Samsung’s earnings have taken a hit because of reduced scale (affecting mostly operating costs per device) and reduced prices (affecting gross margins), so both parts of the profitability equation are suffering at once. Some have suggested that Samsung should go harder after the low end of the market, while others have said that would be stupid because there are no margins there. This is where the above understanding of profitability becomes relevant: yes, the gross margins are lower at the low end, but there’s massively more scale there, which can help operating margins. So it’s not as simple as saying that Samsung should steer clear of the low end because the margins are poor.

Having said that, Samsung does face increasing pressure at the high end, especially since Apple has closed a major competitive window with the new iPhones launched a few weeks ago. So it may be increasingly important for Samsung to focus on the low end. However, that’s easier said than done too: the low end is arguably the main focus of many of the most aggressive moves in the industry at present, with Microsoft providing reference designs for cheap Windows Phones and Google launching the Android One initiative, along with the Moto G and other low-cost smartphones. It’s not necessarily going to be any easier for Samsung to compete at the low end than at the high end, especially if it wants to generate the kinds of margins it has in the past in smartphones. But there’s not really anywhere else to go: the mid-market is rapidly disappearing as the market bifurcates between premium and low-cost, with very little in-between.

Hence, perhaps, Samsung’s renewed investment in the chip business, as a potential supplier to other players. There’s far less competition in that business than in smartphones, and Samsung already has a very strong role in this business as a supplier to Apple and itself. Even as Apple’s reliance on Samsung chips wanes, there are plenty of other opportunities for Samsung to go after here. But of course the average selling price of a chip is far lower than the average selling price of smartphones. Even if the margins are good, it’s extremely unlikely that Samsung will be able to make up the difference on chips alone. Which means it has to continue to go after the smartphone market aggressively too, even though it faces falling margins and shipments going forward.

BlackBerry earnings: progress on several fronts

Having not written about BlackBerry for months, I’m now doing two posts in one week! Normal service will resume shortly. But I did want to quickly cover BlackBerry’s earnings today, because as usual many of the people covering them are misunderstanding what’s happening and focusing on the headlines instead of the underlying trends.

First off, BlackBerry’s results look horrible on the face of them. It’s losing money, it’s shrinking, it’s hardly selling any devices, and so on and so forth. If you compare them to almost any other handset vendor out there, they come off looking pretty bad. But looking at BlackBerry as just another handset vendor is making the very mistake I warned against in my post earlier in the week. BlackBerry’s future involves devices, to be sure, but it goes well beyond them. So here’s a quick take on what I see by way of underlying trends at BlackBerry.

Device shipments

Since so many people are fixated on device shipments, let’s start there. The headline here is that shipments fell significantly year on year, which is usually the best comparison to make to avoid focusing on seasonal trends. However, when a company is in turnaround mode, it’s worth looking at quarter on quarter trends too. In addition, the real number to focus on is sell through and not shipments, because that reflects what the company is actually selling without the effect of reductions in inventory. The chart below shows three key metrics related to the company’s device sales for the last few quarters:

BlackBerry device salesTaking each of those lines in turn: Continue reading