Live blogging Feed

Tuesday, September 20, 2011

Growth and punishment: The vector space model

Growth and punishment: The vector space model:

After processing more than 1500 data points on the performance of thirteen technology companies, patterns are beginning to emerge. The steps so far:

The final step is to plot the changes in the relationship between pre- and post-crisis for the set of companies normalized to the same starting point and then classifying them:

The chart shows how the “average P/Es” changed after 9/30/2008 vs. how the companies performed during those periods. An evocative categorization is suggested for the four quadrants.

One way to read the data would be as a degree of effect of the crisis. For example, in the case of HP, its P/E dropped by about 6.5 while its average growth decreased by 41 basis points giving an average of 0.16 drop in P/E for every basis point reduction in growth. That would be a fairly modest impact. In the case of HTC, the impact would be even less. The effect on P/E was zero even though growth dropped by about 78 basis points. At the same time the impact on Microsoft was severe with the P/E dropping as much as for HP even though its growth only moderated by 5 basis points.

Therefore the “slope” (or angle in vector parlance) of each of the lines in the chart above may be interpreted as the severity of discount in value that the crisis invoked. Those values are shown the following chart:

A strong bar on the right half of the chart implies severe impact, a bar around the middle implies minimal impact and a bar in the left represents an inverse relationship where the market priced in opposition to performance[1]. What should be considered along with this chart is the quality of earnings and consistency of performance (which are shown in the posts linked above.)


As noted in a previous article, there are really only six companies which had strong separation between pre- and post-crisis valuations: Apple, HP, RIM, HTC, Microsoft and Google. They are also the companies which had strong P/E drops (except for HTC). As a group they are also the companies with the best growth stories historically. This is perhaps the reason for the drop. Basically, because these were strong companies they “had the most to lose”. The market “punished” them more severely than the weaker companies because the weak had already been discounted.

The other observation is that Apple and RIM are not treated particularly differently. The drops in P/E are at about 1 for every basis point drop in growth. This means that the the degree of value lost is about the same as a function of growth loss. Microsoft seems to have been the most affected. We can argue that it is perhaps weaker than the others strategically so this is reflected here, but I doubt it’s weaker than RIM. Microsoft is also in contrast with Google which seems to have been given a lot of benefit of the doubt.

Nonetheless, this framework offers some hints on how technology companies have been treated after the crisis. There is no rightness or wrongness about this treatment but it may indicate potential for significant reaction in the opposite direction if and when macroeconomic conditions improve. These vectors are, in a way, proxies for volatility or “beta” in terms of correlation to the overall market. The greater the beta, the more amplification to market movement; the greater the slope value

As always, Apple is the canary in the coal mine as far as sentiment is concerned. What should be watched is the pricing reaction as Apple moves from a net reduction in growth to a net increase in growth. How will P/E react? The argument on this site so far has been that growth is exceptionally strong for Apple and it’s not being reflected in the price. However, in this analysis we’ve stretched the time frame and did a longer retrospective. On this new time scale growth is not yet at the levels Apple enjoyed in the heady days of 2005/6. Perhaps this growth will return and when it does we’ll get to see which way Apple’s vector moves.



  1. An inverse relationship with value can happen for various logical reasons, for example that a company has volatile performance which is depending on anomalous events.

Biggest mobile loser? The non-smart phone

Biggest mobile loser? The non-smart phone:

Yesterday comScore published survey results for EU5 (France, Germany, Italy, Spain, UK) on smartphone use and installed base. The headline is very similar to what would be written about the US: Android had phenomenal growth over the last twelve months. I also noted that the apparent growth of Google (16.2% share change) seemed to be matched by an apparent decline of Symbian (-16.1% share change.) However the reading of the data is not so simple.

In order to understand what has happened to usage, it’s much more valuable to look at consumption and the actual number of users rather than change in share of a subset of the market. Consider the following charts:

The bar chart shows that three platforms gained users in the last year (ending July) and that there were negligible losses in users by others. Negligible except for non-smart phones. A remarkable 29 million users moved from non-smart phones to smartphones. Those switchers switched into Android (16 million), iPhones (6.3 million) and RIM (3.4 million). Symbian even gained about 330k users. Microsoft lost about 1 million but its platform has not been actively promoted in that time.

When seen as a pie chart, you can see how the usage shifted from non-smart to smart even more clearly. It does not appear that Android took share from Symbian at all. Android took share from non-smartphones. The table in the link from comScore leads one to conclude otherwise.

I took the same approach to look at the US data (also from comScore). It shows a similar pattern.

In the US, RIM was a net usage loser but its losses also paled to the losses of non-smart devices. In comparing with EU5 the pattern is very similar.

I then combined the EU5 and US data and provided a view into the combined markets (note scale change from previous two charts.)

This shows the change in platform phones more or less being from non-smart to Android and iPhones. The total population is 470 million people–not insignificant. The total number of non-smart devices abandoned is nearly 60 million in one year, which equals the total number of new smartphone users. That’s 12% of most of the developed world switching in one year. If this keeps up, the “tipping point” when smartphones will outnumber non-smart phones in use in these large markets will be in another year.

In terms of sales rate, the tipping point has already happened. Sales of smartphones crossed over in the US and Europe several quarters ago (November 2010 in EU5) and the trend is accelerating. Furthermore, if patterns of mobile technology adoption repeat as they have for the last two decades, emerging markets will follow in two to three years.

This tipping point observation has been repeated on this site a number of times when reflecting on the data in the US. Now it’s becoming clear that the same pattern has been underway in Europe as well.

The big picture from the survey data is just how vast the demand for smartphones is. Some platforms are benefiting more than others in terms of share of growth, but the biggest mobile loser is clearly the non-smart phone. Not only is it an unprofitable product for almost every vendor, it is also being increasingly shunned by buyers.

At the same time, at 450 million users we’re still only looking at less than 10% of the market. As this pattern spreads globally the platform install bases will be measured in the billions of users.

Mobile Impossible

Mobile Impossible:

In yesterday’s post about the “biggest mobile loser” I covered the exodus of users from non-smart devices in the US and EU5. I also said that what happens in those regions tends to happen in other regions with a time shift. In some regions it happens quicker but in most it happens more slowly.

But can we be sure that there isn’t vast non-smartphone growth in other regions? Well, no, we can’t be sure. At least not without access to reliable data.

But what we can track is the overall non-smart phone market and compare it to the smartphone market. Here are the growth rates of the two sub-markets:

The difference is plain to see. We can also note that the non-smart market may be heading into a contraction–something noted by some analysts close to the market–but no real sign of that happening in smartphones.

Beside growth, we can also see actuals and the split of various vendors’ volumes in the market.

It’s also clear that the top line has not moved much in the last few years. Next, the shares of the vendors.

Finally, the before-and-after share pies.

We can see that the growth story, if there is one, is coming from low-end entrants. In late 2008 the main brands had 90% share. They now have about 60% (in a flat market.) The growth is all from “Other”.

What seems to be happening is that buyers in mature markets are swapping (or upgrading) from non-smart to smart and adopting new smartphone brands. At the same time, new users in emerging markets are increasingly adopting new non-smartphone brands.

So the question for the incumbent vendors is extremely important: Facing them are low-end and smartphone entrants engaged in a pincer movement, with obvious success; can the old brands survive?

The answer, as always, depends on whether the competition is symmetric or not. Low end competitors are usually structured around low cost structures. High-end competitors are exploiting new markets and new business models. For an incumbent to survive and fight on both fronts they’d have to adopt both of these models for themselves, simultaneously and quickly.

We can see that Sony-Ericsson and Motorola (Google) have thrown in the towel on the low end. The question of dual-front strategy is mainly relevant only for Nokia, Samsung and LG. LG is on the ropes and may actually be ready to exit the market so that really means Nokia and Samsung. Nokia has signaled continuing interest in the non-smart market (citing “the next billion users”). Samsung also seems to be hanging in there. Can they do it?

I cannot stress how difficult this is or how rarely it has been achieved. The two approaches are if not opposites then at least orthogonal. It’s like Mission Impossible. You can’t be a low cost global vendor while investing in R&D for differentiated platform-based devices. These are shaping up to be businesses requiring completely different skill sets. Few have the ability to host these skills inside one company. I can think only of one example.

RIM and the lamentation of the analyst

RIM and the lamentation of the analyst:

RIM shipped 10.6 million Blackberries and 200,000 PlayBooks in the last quarter. Management noted that their sell-through was significantly higher for Blackberry (13.7 million) but seems to be very weak for PlayBook as the prior quarter saw 500k units shipped. Additional PlayBook units this quarter probably mostly went into new channels in Asia and there were no additional sales into North America or Europe.

The figures for units are very poor. How poor depends on the frame of reference. Consider the shipment chart below:

In terms of the competition, 10.6 million units is less than half what Apple or Samsung sold in its prior quarter. It’s also less than what HTC sold. RIM’s volume rank will likely go to fifth place as a smartphone vendor.

In terms of its performance relative to its own history, the Blackberry volume dropped by 11% year-on-year and 18% sequentially. This is the second quarter that shipments shrank.

In terms of market share, we’ll have to wait for the competitor data over the next six weeks but the market has been growing at an average of 77% for four quarters so any continuation of this trend would imply RIM’s share dropping to single digits.

Finally, the biggest shock has been the decline in profitability. It seems that operating margin dropped from 21% to 13%. The company did incur some one-time charges for recent layoffs, but even without that charge, the margins would be around 16%. This is most alarming. The reason for such drops is that as volumes decrease fixed costs don’t decrease as rapidly or at all. The company still needs to keep sales, administration and engineering staff around and they become a larger part of the operating expenses (vs. the component costs which vary with volume of goods sold).

This effect is well understood by financial analysts and the stock price shows it with a huge drop.

But stepping back to look at the picture above, there is a clear turning point in the company. You can see the elbow in the curve for volumes whose effect is felt so deeply. What’s curious is why the pivot occurred when it did. We can point the finger to competition as the cause. But why was there no effect in the company’s fortunes when the competition actually emerged. It’s been years since the iPhone and even Android entered the market. Yet we see an impact on Nokia and RIM individually at apparently arbitrary points of time.

This is the lament of the analyst: you can clearly and accurately state what will happen but when remains a mystery. It’s the elasticity between obvious causes and their effects that makes this an inexact science or not a science at all. In retrospect, you can say that Nokia’s pivot was triggered by its public execution of Symbian, but that assumes that it was preventable–which we know is not the case. But what caused RIM’s change of growth, exactly? Why did it happen this past spring? Why didn’t the company volumes begin to decline as iPhone and Android boomed in 2009 or 2010? For quite some time RIM seemed immune to competitive pressure. We all were made fools as we called its imminent demise. Then, as Steve Jobs would say, boom!


Piecing together RIM’s performance is becoming more difficult each quarter. The data being presented is increasingly obfuscated by irrelevant detail while major information is omitted. This last management presentation was full of holes, namely:

  • We have no idea of device pricing. Valiant efforts have to be made to piece together that aspect of the business. Matt Richman does a good job but it still requires some guesswork regarding PlayBook pricing to back out Blackberry pricing.

  • Management detailed where Blackberries were selling-out but only detailed sell-in for the PlayBook. Obvious spin.

  • They are cherry picking the data they present and each quarter it’s a different story making pattern recognition impossible.

This leads to an erosion of trust. Observers are faced with the problem of increasingly guessing what is happening each quarter. For instance, regarding pricing I prefer to include service revenues in the analysis of device sales because device+service is what is being bought by the user and vendors which can offer services as part of the device get a justifiably higher value for the product–value that I think needs to be considered as advantageous. Nokia and Apple also account for services (though not apps) as part of device sales and it makes it more convenient to compare these businesses.

Given RIM’s smokescreen, this quarter I decided to stop trying to guess Blackberry pricing and used Operating Income/Units sold as the average selling price.

The tell-tale signs suggesting a platform’s demise

The tell-tale signs suggesting a platform’s demise:

In the post on OS turning circles, I used the concept of a radius of turning as an analogy for agility. One problem with the analogy is that turning in circles implies a return to a starting point or at least a closing of the loop. The idea is that there is lifecycle repetition. However, in reality, this does not apply to the world of operating systems.

An OS, as a platform, usually has a finite life. It is born, grows and usually reaches a point where it is no longer supported. Sometimes, a new platform is born to take its place from the original owner but more often a replacement comes from a new challenger company.

So rather than circles, the analogy of OS lifetimes may be more accurate.

If we do think of platforms as finite, then the natural question is what causes an end? We need to look for patterns which may indicate when a platform is reaching end of life.

The difference in this analysis is that the measure of “age” of a platform I use is not time per se but versioning. The logic is that each major version is a meaningful and significant improvement in a platform which needs to be delineated, marketed and celebrated. It embodies the business logic as well as the engineering logic of the platform custodian.

Taking the data from the last post I added a few more platforms: Symbian[1], PalmOS and Blackberry OS[2] to seek out patterns. I also separated the desktop/portable OS’s from Mobile OS’s and plotted these version-demarcated lifespans.

One thing to observe is that the scales of the two charts are comparable. There are examples of short- and long-term version updates and the number of iterations (lifespan) can be similar. I noted also that there were several platforms which have reached end-of-life[3][4]. Those platforms have a big black dot at the end of the line.

That leads to another observation. The end of a platform seems to be indicated not by simple age (the shortest lived was six years and the longest lived was 16) nor by the number of versions (PalmOS lived for five while MacOS lived for nine). Instead the end of life is most clearly visible as a lengthening of the development cycle.

Note that each platform that ended was preceded by a spike into the vertical–a significant delay in the release of a version. The data is one thing, but it’s anecdotally supported by observation. Industry observers note that delays in improving the product are symptoms of some fundamental architectural or marketing roadblock. In the case of Mac OS, Apple struggled to bring modernity to its “Classic” OS. It need memory management, more reliability and a better file system to support the move to networking and media hub use that defined the consumer expectations of a PC.

There was a change in the basis of competition, away from pure productivity and more toward entertainment that turned out to be more demanding in new ways. Apple had to move away from Mac OS and lost time with its internal Copland effort before punting with NextSTEP. Similar transitions are visible with Palm (from PalmOS to WebOS), Microsoft (from Windows Mobile to Windows Phone) and Nokia (from Symbian to MeeGo) and RIM (from Blackberry OS to QNX). In fact, survival of a transition is relatively rare and never without significant pain and loss of value or share.

This is also understandable through the lens of disruption theory. As a product reached the point of being good enough, “breakthroughs” are harder to come by. Engineers and marketers struggle to push the product into increasingly rarefied strata of performance. The old architecture does not fit the new demands but it’s crammed into them anyway. This last big push is then followed by a stall and ultimate demise. Meanwhile, an entrant gains lift in the rich atmosphere of new bases of competition with an architecture that’s built specifically for it. The process then repeats.

And so the charts relate the same story of sustaining improvements followed by inevitable last gasps that Clayton Christensen first illustrated[5] in The Innovator’s Dilemma. A book that came out just as the first mobile platforms cataloged in these charts began their ascent.


  1. Versions of Symbian do not match easily to integer values. I used the versions as recorded by The sequence is as follows: EPOC Release 3: “1″, Release 4: “2″, Release 5: “3″, Symbian OS 6.0: “4″, OS 7.0: “5″, OS 8.0: “6″, OS 9.1: “7″ (9.0 was deproductized), Symbian ^3: “8″. Symbian ^4 has been cancelled. These original version numbers are noted on the chart.

  2. The times recorded are for “general availability of product” which in the case of mobile OS’s means the time when a device using that OS was released.


  3. End of life is defined as the last version generally available. In some cases (e.g. PalmOS 6) newer versions are built but they may not be released into a working complete product.

  4. Windows Mobile is treated differently here than in the previous post. I chose to declare it EOL after version 6 and consider Windows Phone as a separate OS. This is because the name change is indicative of a break with the past. As before, I defer decisions about continuity to the developers and/or marketers who choose the naming conventions.

  5. See slides 3 and 5 here

Monday, September 12, 2011

Growth profiles of 15 companies

Growth profiles of 15 companies:

The previous article showing the profile of Apple’s growth vs. its P/E prompted a similar review of a set of comparable companies. The cohort is composed of:














We made one change to the growth data where the Net Income growth is not quarterly year-on-year but average of four quarters year-on-year. This reflects the fact that P/E is also a trailing twelve months’ earnings. It also has the benefit of smoothing the growth data making it easier to discern.

Here are the charts:

Note that all charts are to the same scale (P/E from zero to 50 and Net Income growth from -100% to 200%). Some clipping of the values is possible. Growth from unprofitable periods are undefined (i.e. when measuring growth from a to b, if a is negative, the data is omitted.) Growth from a positive period to a negative period is defined as -100% (i.e. when measuring growth from a to b, if b is negative, the value is defined as -1).

There are some anomalies, for example growth due to an acquisition or loss of profitability due to re-organizations, however the patterns are probably what matter most.

Next post will discuss the relationship between P/E and growth in the pre- and post-crisis time windows for all these companies.

August browser stats: Safari dominates mobile browsing

August browser stats: Safari dominates mobile browsing:

Our browser stats post is late this month. The source we normally use, Net Market Share, has changed the way it reports its data. This is good and bad. Mostly good, but it took extra time to retrieve the data and then decide what to do with it.

The good part is that we now have separate statistics for mobile browsers and desktop browsers. This answers long-standing demands to break this information out to take a closer look at that small but increasingly important market. The bad part is that the new figures are much harder to compare to historic ones; Net Market Share has completely separated mobile usage from desktop usage.

Read the rest of this article...

Read the comments on this post

Sunday, September 04, 2011

The proliferation of mobile platforms

The proliferation of mobile platforms:

While some mobile platforms are being withdrawn from the market, others are being introduced. The net is that there are more mobile platforms announced for 2012 than ever before. The following chart shows the lifespan of the platforms that I can recall.

Amazon’s entry is only rumored but I think the description of it is detailed enough to be credible. The day after the scoop with Amazon, Baidu announced the Yi platform. Baidu is the sixth most visited site in the world, so it’s not a bit player and it makes as much sense for Baidu to have a mobile platform as it does for Google.

Aliyun is Alibaba’s latest and it, like Amazon’s, is positioned on commerce and shopping. I also included the platform underlying Barnes & Noble’s Nook for consistency.

Many of the newcomers are Android-like or Android-derived platforms but cannot be rightly called Android as they are not licensing the Android trademark nor any of the proprietary part of Android. They may run some subset of Android apps but that is not guaranteed to continue.

It’s also worth nothing that many of the newer Android-derived platforms are being launched by Google’s competitors and indeed seem to be designed to take value away from Google.

The epic struggle Android vs. Google is sure to continue.

Friday, September 02, 2011

The third ecosystem. What are the odds?

The third ecosystem. What are the odds?:

The latest survey from ComScore breaks down the installed base of specific mobile platforms in the US. Tracing the data over time shows the following charts:

In the last month Android gained about 2.9 million users, iPhone 1.3 million, Blackberry lost 530k users and Windows held on with a gain of 132k. Other platforms had a net loss of 95k users.

In the last 12 months, Android gained 25 million users in the US. iPhone gained 9.5 million while Blackberry lost 3.2 million and Microsoft lost 1.6 million. Other platforms had a net loss of 1.2 million.

The total net gain of smartphones was about 29 million new users.

RIM switched from being a consistent net gainer of users to a consistent net loser of users in October 2010. Windows Phone is showing signs of holding the line on user base erosion but share remains below 5% (now at 4.7% vs. 4.6% last month). To put the mountain-sized hurdle in perspective, Android now has 7 times more users in the US while iPhone has about 5 times more. To become the largest mobile platform in the US, as some analysts are predicting, Microsoft has a 12:1 disadvantage that looks to continue to grow.

Those are some pretty tough odds.

Thursday, September 01, 2011

40 Percent of U.S. Mobile Users Own Smartphones; 40 Percent are Android

40 Percent of U.S. Mobile Users Own Smartphones; 40 Percent are Android:

Don Kellogg, Director of Telecom Research and Insights

Forty percent of mobile consumers over 18 in the U.S. now have smartphones, according to July 2011 data from Nielsen. Android is the most popular operating system, with 40 percent of mobile consumers reporting they have a smartphone with an Android OS. Apple’s iOS is in second place, with 28 percent.smartphone-marketshare

Among those who say they are likely to get a new smartphone in the next year, approximately one third say they want their next smartphone to be an iPhone and one third say they want an Android device. However, among those who say they are usually the first to embrace new technologies, “Innovators” or the earliest of early adopters, Android leads as the “Next Desired Operating System” – 40 percent for Android compared to 32 percent for iOS. (Survey respondents were asked several questions to determine their attitudes toward new technologies.)


Among likely smartphone upgraders, it is the “Late Adopters” who are most likely to say they are “not sure” which operating system they’d like in their next smartphone. In politics as in smartphones, these “undecideds” will be the ones device makers will be hoping to win over.

  • For related insights, join us for our free webinar, All About Android on September 15.

Mobile App Inventory Hungry Enough to Eat Internet Display Ad Spend

Mobile App Inventory Hungry Enough to Eat Internet Display Ad Spend:

Smartphones app usage, facilitated by explosive iOS and Android device adoption, has created among the fastest-growing media channels in the history of consumer technology. Flurry estimates that, worldwide, over 600 thousand apps are available for over 350 million iOS and Android devices. On average, consumers have downloaded over 65 apps per device.

While micro-transaction models, largely associated with free-to-play games, have proven the most lucrative business model for iOS and Android apps, there have been big bets placed on advertising. In addition to its own iAd initiative, Apple acquired Quattro, a mobile ad network, for $275 million in January 2010. This was shortly after Google announced its intention to acquire Admob, a rival ad network, for $750 million in November 2009.

In June 2011, Gartner projected that mobile advertising revenue would double to $3.3 billion worldwide in 2011, and grow from around $300 million to over $700 million in 2010 in North America. eMarketer, a research firm, predicts that U.S. mobile ad spending will top $1.1 billion this year.

In this report, Flurry focuses on the size and growth of available advertising inventory within iOS and Android applications. We used data from over 100,000 applications tracked by Flurry to estimate the size of this media channel. The chart below shows that U.S. app inventory is not only growing at a staggering rate, but also poised to absorb the equivalent of the entire U.S. Internet display advertising spend by the end of this year.

USappInventory vs USonlineDisplayAdSpend resized 600

Reviewing the chart, we see that U.S. mobile app inventory has grown aggressively over the last year. With its growth trajectory, it will be able to absorb the entire U.S. online display ad spend by the end of the year. Another way to look at this is that, in approximately two years, mobile app inventory is growing so aggressively that it could easily meet the demand of a mature, 15-year-old form of online advertising.

To arrive at these figures, we first tracked the average number of ads shown per application session, which we found to be 4.3. The average application session is 4.2 minutes. For reference, the average session length of a website is just under 1 minute. We then looked at the number of sessions. Flurry tracks about 20% of all sessions in the market, and so we grew our numbers accordingly to come up with a market size.

We compared this inventory with the net spend on display advertising in the US. The US market currently spends a little over $12bn per annum on online display advertising. We assumed a conservative CPM (cost per 1000 impressions) of $2.50 for mobile application inventory. As a point of reference, a typical 30 second video on a large video streaming website such as Hulu has a CPM of $10-$15.

We at Flurry see four reasons why the market is growing at such a fast rate:

1) Smartphone growth – over a million smartphone devices are currently being activated on a daily basis

2) Publisher growth – The App store now has over 400,000 apps in the market and Android, with over 200,000, is catching up quickly

3) Session use growth - Flurry has previously found that smartphone users now spend more time in mobile apps per day than the average Internet users spends online.

4) Publisher integration of ads – with larger screens, targeting, and increased adoption of mobile applications, more publishers are integrating ads into their apps

Not only is inventory growing, but Flurry has also found that the average user of a smartphone is a very attractive target for advertisers. With a sample of more than 60,000 app users, we used location data and zip code statistics available from the U.S. Census Bureau to understand their demographics. On average, smartphone users are better educated and earn higher household incomes than the average of the U.S. population.

HHI smartphone v USaverage resized 600

Additionally, looking at age and gender, we find that U.S. smartphone app users cluster into younger age groups and trend slightly more female.

Age and Gender Smartphone v USaverage resized 600

In 1994, was the first company to start selling display advertising in large quantities on the Internet. Back then, it took over six years for advertisers to embrace this model. For mobile apps, less than four years into their growth cycle, a critical mass of highly attractive consumers has been achieved. With growing awareness by brands and advertising agencies, we now expect digital advertising on mobile to take off in earnest.