Horace Dediu

A 2 minute video introduction on how I use Perspective to create interactive presentations:

 

We cover three topics:

1. Safety and air travel: the bases of performance that don't get rewarded
2. The unintended consequences of litigation: US v. Microsoft and lessons learned for Apple v. Samsung
3. Product Portfolio theory: is focus impossible for everyone but Apple?

5by5 | The Critical Path #51: Product Portfolio Theory.

The first of a series of padcasts capturing the Asymconf event is ready. This presentation includes a discussion on the formation of Amsterdam and the innovations that both made, and were made by, The Dutch.

It is the story of the Netherlands as seen through a disruptive lens.

The introductory remarks are captured with audio and video mixed with data and content as presented live.

The Story of Amsterdam is free.

View it using the latest version of Perspective on your iPad.

We can put all of our products on the table you’re sitting at. Those products together sell $40 billion per year. No other company can make that claim except perhaps an oil company.

We are the most focused company that I know of, or have read of, or have any knowledge of.

We say no to good ideas every day; we say no to great ideas; to keep the number of things we focus on small in number.

Tim Cook said this in February 2010 at the Goldman Sachs technology conference.

Since then the only product that has been added to the kitchen table has been the iPad. The sales level however, has increased in proportions shown in the chart below.

The revenues are shown with their contributory products and the costs of those products. The payments for costs of sales as well as R&D, SG&A and Taxes are then subtracted revealing the Net Income (in green). This is done for the second quarters of 2010, 2011 and 2012.

Since Tim Cook made the analogy, the  table holding the products has not gotten any bigger but the sales level has more than doubled while profits have nearly tripled.

In his talk he cited revenues of $40 billion (for the pervious year, 2009). In the last twelve months Apple's revenues were $148 billion.

Tim Cook went on:

I think this is so ingrained in our company that this hubris you talk about that happens to companies that are successful and sole role in life is to get bigger, I can tell you the management team at Apple would never let that happen. That's not what we're about. [We have a] Small list of things to focus on.

The irony is that by thinking small, Apple became the biggest company that ever was.

The trick to thinking small is the same as that of "focusing" which means saying no, a lot. I touched on this in February 2011. Focusing is the willful rejection of the Nobel-prize winning theory of portfolios.

From the wiki link above,

Diversification eliminates non-systematic risk, but at the cost of increasing the systematic risk. Diversification forces the portfolio manager to invest in assets without analyzing their fundamentals, solely for the benefit of eliminating the portfolio’s non-systematic risk (the CAPM assumes investment in all available assets). This artificially increased demand pushes up the price of assets that, when analyzed individually, would be of little fundamental value. The result is that the whole portfolio becomes more expensive and, as a result, the probability of a positive return decreases (i.e. the risk of the portfolio increases).

This theoretical view into asset management applies in spades to product portfolio management. The idea that it's better to spread one's bets and cover all segments is "safer" has been thoroughly debunked by Apple.

Understanding the "fundamentals" of a product is far more important than having lots of products for the sake of diversification alone.

The fundamentals of a product are knowing its job to be done and thus its requirements which, when well executed, will position it precisely and unambiguously on an opportunity.

Google, Microsoft and Apple all had reasonably good second quarters. Microsoft's revenues grew by 4%, Google's by about 22% and Apple by 23%. Apple's growth was disappointing but the other companies weren't. The revenues for the three companies are shown in the following charts:

 

I also showed the performance of each company's operating income in the bottom three charts.

Microsoft did declare a one-time write-off of $6.2 billion which it spent on aQuantive a few years ago. That was booked as contributing to a loss of $6.67 billion for its Online Services division. Without that one-time charge consolidate operating income would have grown at about 3.6%. Google's operating income grew at a rate of 11% and Apple grew (again, disappointingly) at 23.4%.

Apple's operating margin fell to 33.1% while Google's fell to 29.2%. Microsoft's margin evaporated due to the one-time charge. If we back that charge out, the margin would have slightly topped 35%.

Comparing the contribution (in operating income) for each division or product line (where possible) gives the following view:

I tried to remove some noise by isolating the iPhone and iPad and consolidating Microsoft overall and Google.

By also smoothing out seasonality with a four-period moving average we can observe the pattern of growth for the iPhone and iPad relative to the other companies' overall performance.

The crucial question remains if and when the iPad will be a bigger business than Google and/or Microsoft. The iPhone climbed over both some time ago.

``It’s costly and has unpredictable consequences. That makes it a very bad business model. I suppose both companies have agendas that are not visible in court and perhaps Apple is signaling to other parties, and perhaps Samsung is using it to raise its profile. It still seems that unintended consequences may arise and the result could be very bad for everybody,’’ said Horace Dediu, founder of equity-research firm Asymco, in an e-mail interview with The Korea Times.

The above was a quote from the article  Samsung can’t afford to lose Apple fight by Yoo-chul Kim from The Korea Times. Yoo-chul contacted me via email with a series of questions. I want to share all his questions and my answers as well:

Samsung wides market gap with Apple, according to research firms. You think the gap will further be widen and please tell me why

As Samsung converts its phone portfolio to be all smartphones, it is very likely to reach much higher smartphone volumes. One year ago Samsung sold about 56 million feature phones. Last quarter it sold about 43 million. The smartphone portion increased from 20 million to 50 million. At the same time, Apple has maintained a growth in its smartphone business of about 100% on a yearly basis. I expect that to continue into 2013 as well. Therefore it's possible that Samsung could sell 290 million smartphones in the next 12 months and Apple could sell about 200 million in the same time frame.

However, it's important to understand that the companies compete on different bases. What I mean is that what generates a profit (i.e. payment by the consumer above what a product costs) is different for each company. Apple is a platform company that sells access to the iOS ecosystem across many other device types. Samsung sells consumer electronics with limited "network effects" but desirable features. In this view, the gap in device sales is "asymmetric" and not reflective of the actual competition which is between Google's Android and Apple's iOS as forms of access.

Samsung's rapid rises mainly due to the company's heady dependence
on Google Android. If Google applies better software for Motorola,
then you think Samsung will face difficulties?

Think of Google as a supplier to Samsung. No manufacturer wants to have a situation where they are dependent on only one supplier, especially when there are huge substitution costs for another supplier. And even more especially if that supplier owns one of your competitors. However, it's unlikely that Motorola will become a strong competitor to Samsung. I also don't expect Google will damage their OEM "customers" by favoring Motorola. Motorola can't become big enough quickly enough to be a threat to Samsung. What is more likely to be a problem for Samsung is ZTE and Huawei. They are on the same trajectory that Samsung was on 10 years ago. For Samsung to beat them in the future it will need to develop products that nobody can copy.

How about software? Samsung is good to roll out price competitive
models, however, software is always problem. Any advice for this
issue?

Software is hard and expertise can't be quickly acquired. It takes time. Only advice [I have] is that efforts must be patient and there need to be many experiments like Bada and Tizen. Google started with Android in 2005 and Apple started with iOS in 2003 or 2004.

How about the patent dispute with Apple and Samsung. The patent war
is helping Samsung see a rise in its brand awareness and that also
worked for Samsung to grab more shares in smartphones?

It's very difficult to measure the impact of litigation, both in the positive and negative. It's costly and has unpredictable consequences. That makes it a very bad "business model". I suppose both companies have agendas that are not visible in court and, perhaps, Apple is signaling to other parties and perhaps Samsung is using it to raise its profile. It still seems that unintended consequences may arise and the result could be very bad for everybody.

Samsung wants to become a market creator not just a fast follower.
For me, it's quite negative for Samsung to create a wholly-new
eco-systems because Samsung's edges are coming from manufacturing, not creative thinking. What do you think? If Samsung wants to be a
creator, what conditions should Samsung need and what things should
the company do?

Unfortunately, manufacturing excellence is not enough to remain successful long term. US companies were successful, then the Japanese, Taiwanese, Hong Kong, and now Chinese. Companies either adapt or become irrelevant. Building ecosystems is the new basis of competition but it's not the only option. Unfortunately we don't know what will be necessary to build the next basis. What matters is experimentation and the building of new businesses. I can't offer specific recommendations as there is no recipe book for how to create new markets.

For your views, who is much better so far since the trial has begun
in the US. Samsung or Apple?

I cannot comment on this. Performance in court can only be judged by those present and I'm sure everybody present disagrees.

You think Apple is enjoying its so-called a homeground advantage
for the trial and is there any possibility that the issue to go for
the highest court in the US. My Samsung contacts say that's the
possibility, though.

Again, I don't have an opinion on legal matters. Broadly speaking, IP litigation is similar to playing the lottery.

Need your perspectives about Google's support for Samsung. Google
confirms its supports for Samsung for an united front against Apple
because the fight between Samsung and Apple is a kind of version that
represents Google and Apple fight.
Any comments about that?

I believe Google does support Samsung. I question however how this relationship will develop over time. I attached below commentary from a researcher at Deutsche Bank on Android in China. I think it has some clues about how Google treats its partners.

What are pre-conditions before Samsung making an action to sign a
peace treaty with Apple?

It depends on the motives behind Apple's actions. I think they want to see a "no copying" rule. They don't mind competition and Android could be a help in the long run by moving users from feature phones to basic smartphones which could later be upgraded to Apple phones. But what I think they want is to signal to others that direct copying is unacceptable. Essentially, the work of their designers is very important to them and copying destroys [the designers'] incentives to do breakthrough products.

---

See also: Who Cares If Samsung Copied Apple? by James Allworth - Harvard Business Review

One of the remarkable patterns associated with the iPad has been its growth rate. Looking at the ramp of the product relative to the iPhone and iPod touch reflects this sense of raised expectations:

 

At the same point in time, and adjusting for seasonality the iPad is outselling the iPhone by a factor of 3 and the iPod touch by a factor of about 5. This is also reflected in the installed base of devices, where the 84 million iPads sold through its first nine quarters is more than three times the iPhone base at a similar point in time.

Of course, the world of 2012 is different than the one from 2009 when 25 million iPhone sold. iPhones are available in far more points of sale and to many more buyers world-wide. But the interesting contrast is between the iPad and the iPod touch.

I've always maintained that the dynamic difference between the iPad and the iPhone is that it's unrestrained (and unsupported) by the operator channel and all that it implies. The iPad is a pure computing product whose means of communication is unregulated. That means transparency in pricing and transparency in distribution.

But so is the iPod touch.

The iPod touch is a "nano iPad" (or as detractors put it, the converse of that.) So shouldn't the iPod touch behave as the iPad?

We don't know exactly how the iPod touch behaves. We don't have specific figures on its sales. What you see in the charts above are my estimates for the iPod touch based on company statements that "about/above 50%" of iPods are touch versions. That and the cumulative sales totals that are occasionally reported.

However, we have received an unexpected morsel of data from the Samsung v. Apple trial that may help characterize the two products: US sales data. Having the US data and knowing/estimating the global data gives us an estimate of the international (ex-US) sales for all three products.

I indexed the three products to measuring quarters since launch. The pattern here is that the iPad diverges from the iPod touch in its international appeal. Since the iPad expanded internationally even faster than the iPhone, the issue is not only one of rapidity in appeal but also a qualitative difference.  The iPod touch seems to have stabilized at 50% non-US sales while the iPhone is between 60% and 80% outside the US.

The iPad reached 60% non-US sales in its first year. Its last quarter was 70% outside the US.

So what's going on?

The characteristic value of the iPad is in its appeal to "computing non-consumers". Computing has come to mean media consumption, education, communication and entertainment. These jobs to be done are casually done with much more ease on an iPad. The iPod touch can do these jobs as well, but not as comfortably. The iPod touch is "absorbable" by younger consumers and typically they have their hiring done by their parents. Those parents are less likely to indulge where incomes are lower.

But the universal appeal of the iPad is evident in its global visibility. Like the tasks it performs, the iPad itself is becoming less of a mystery.

ifoAppleStore's Gary Allen alleges that John Browett, head of Apple Retail, felt the stores were "too bloated". He cites "numerous tipsters" that Browett ordered a reduction in the number of employees. As the information on employees and visitors is public, we can quickly test the assumption.

The following graph shows the relationship between Retail employees and Visitors on a global basis.

The total figure for the last quarter was 41,000 employees for 83 million visitors. This is above the expected ratio as shown by the line shown in the graph. Reversion to that line would imply that about 35,000 employees should have sufficed during that quarter.

However, Apple's business is cyclical and employment is not--or at least not on the same frequency of cycle.

During a slow, transition quarter, there will be over-employment and during a rush holiday period there would be under-employment. Consider the right-most point in the chart. It represents Q4 2011 when there were 110 million visitors and only 800 more personnel than during the last quarter.

Overall, this ratio of visitors per employee has been relatively steady, though declining.

Perhaps the company should check this decline. Perhaps the company has not been able to match demand with supply as well during the last six months. Perhaps they are overstaffed. However, I doubt that there has been a decision to reduce the overall level of service offered. Or that service is not paramount.

The accusation that employment is being sacrificed for the sake of profits seems disingenuous. The stores were never designed as a profit center. They have to be profitable in the sense that they have to justify themselves. But we need to recognize that they rely on a brand and product mix that is unique to that store. In return, they are designed to sustain that brand and offer a service rather than merely a sales channel.

For the head of the division to misidentify this fundamental purpose for the operation would be catastrophic and I would expect a lot more evidence to support such an extraordinary claim.

The crumbs of data falling off the Samsung v. Apple trial table get some scrutiny. Horace expands on some of the hints from the partial release of information and then continues with a discussion of how market data is collected and whether it should be trusted. That leads to a question of whether private (or paid) analysis is "better" than public (and unpaid). The benefits of having access to the vastness of collaborators online and the public sources of info might be tipping the balance. Finally, we talk about how big ideas go from sounding impossible to being inevitable and who gets rewarded for making them so.

via 5by5 | The Critical Path #50: From Impossible to Inevitable.

The second quarter showed continuing growth for Android with 19 points of share growth from a year ago. Only Windows Phone showed a gain y/y in share up 1.6 points to about 3%. In the same time, Bada lost half a point, iOS lost 2.1 points, RIM lost 7.2, Symbian 11.4.

The result is shown in the graph below:

I also show a "before-and-after" pair of pie charts which show the difference in smartphone platform market share from the same quarter three years ago. Android went from 3% share to 67% while Symbian went from 41% to 4% and RIM went from 19% to 5%. iOS increased its share from 13% to 17%.

If this trend continues Android could reach 50% of all phones sold in a few years. That's a run rate of 200 million per quarter. And it's likely that 100 million of those will be shipped by Chinese vendors.

However, looking at the global picture can also miss some important trends in local markets. Had one looked at the global smartphone business in 2005, the conclusion would have been that Symbian would be the dominant platform in perpetuity.

What affected the global market was a change in the US: the launch of iOS and thereafter of Android. The US, a smartphone backwater until then, became the most vibrant market and today it's the most penetrated and has the highest ratio of smart to non-smart of any of the major regions.

The curious thing about this vast market is that it's still not truly open. There are technology and political reasons why some platforms are permitted to sell in one region or not. The iPhone is not available on either the largest Chinese operator or on the largest Japanese one. It's largely absent from India because of distribution (i.e. subsidy) and much of South America. Even in the US, the iPhone is still, after five years, not available on one of the top four operators and has only reached another less than a year ago.

What intrigues me is what would the market look like if distribution were not an issue. What if there was equal choice to all buyers. We can't test this question directly but we can ask. But there is some pattern we can look at in the US data. Thanks to recent revelations at the Samsung vs. Apple trial we can see the effect of increasing distribution in the US.

The data on smartphone market share by carrier can be blended with data from carriers about iPhone activations (which are a proxy for sales, see note below) to give a more precise figure of market share in markets where iPhone participates.

For example, when the iPhone was only available on AT&T, one could measure iPhone market share vs. the global smartphone market, or the US smartphone market or within AT&T alone. Some assumed that the AT&T share would be reduced once the product would become available on Verizon or other US carriers. However there was no appreciable drop. Share in Q4 2010 was 63% at AT&T and it was 63% a year later after the iPhone was available for four quarters at Verizon.

So the share of iPhone within those operators which carry it tells a story independent of distribution.

The following graph shows the share of the iPhone vs. Samsung in the market defined by its availability. (I.e. the volumes of smartphones sold by AT&T throughout the time period, Verizon Q1 2011 to end and Sprint in Q4 '11).

The share of iPhone thus defined shows a slight decline but it's an overall number that is remarkably resilient as distribution increases and is remarkably high overall. Note that globally, the iPhone gets only about 17% share of smartphones--a value that has not varied greatly since mid 2009.

So at least in the US, on carriers where it's available, the iPhone is holding on to a significant share. In a global market it's expanding distribution but not yet covering the world. As a consequence it's keeping up with overall market growth and staying between 17% and 20% share.

The big question is what will happen if and when its availability matches that of Android in terms of points of sale. Or even that of BlackBerry or Symbian. Is America a different market from the rest of the world? It was different in being isolated while Symbian roamed globally. It was different in being a launch platform for BlackBerry and then iOS and then Android. But now that iPhone is beginning to be widely available there, will it be exceptional in how it rewards Apple?

---

Note:

1. Activations data can be seen next to Apple's testimony of US sales in the following graph. Note that more phones were sold by Apple than were activated in total of the life of the product. 86 million sold and 80 million activated. This implies that an analysis based on activations undercounts the iPhone. 

50.2 million.

That's the estimate from IDC. So why bother asking?

Because that is an estimate. Of the 104 million Android phones shipped in the quarter (itself an estimate from another, possibly different methodology), I could only account for 7 million actually reported. That figure comes from a close reading of an investor presentation from Sony. HTC does not report their shipment numbers. It stopped some time late last year. Neither does Motorola now that it's a part of Google. Huawei is silent except for setting targets and ZTE published a press release citing IDC's estimate of its own shipments.

But most glaring of all is the absence of any mention by Samsung of its performance. The company stopped reporting any data on either overall phone shipments or of smartphones within that total since Q3 2011. We may have been able to estimate Samsung if we had more competitor actuals, allowing us to back into a figure. But we don't.[1]

Which leaves us with IDC's estimate[2]. But the problem is that since the industry is growing so quickly it is very sensitive to assumptions. Consider how difficult it is for consensus estimates for Apple's iPhone shipments to come near the actuals--and that's for one quarter, and knowing all the previous quarters with precision. The absence of visibility into the assumptions made by market analysts (or their methods) should lower confidence in the results.

Consider that IDC specifies 50.2 million Samsung smartphones, implying an accuracy down to 100,000 units. Is this accuracy believable?

There is reason for doubt.

I start with the following graph:

It shows the estimates for Samsung global shipments (latter four quarters are estimates without the benefit of company reports.) I separated the component of shipments that Samsung reported as part of its submission to the Samsung v. Apple trial in California. Note that this (orange) segment consists of most US smartphone shipments. It's not all US Samsung shipments. It notably excludes the Note, the latest Nexus and the Galaxy SIII.

Without these "non-accused" phones, it would appear that Samsung's US volumes dropped to 4% of global totals from a 10% average[3]. In order to return to 10%, the US totals would need to be adjusted as shown below:

 

I added 1.5 million units in Q1 and 4 million units in Q2 (to account for Galaxy SIII).

So far this seems consistent with the total estimated global shipments of 50.2 million from IDC. US units would result in 12% of global shipments.

However, there is more data we can test against. First, we are also given pricing for the US phone portfolio, or at least we have revenue per unit sold. This allows us to get an idea of what revenues can be obtained for the smartphones in the US, and possibly world-wide[4]. Knowing prices for smartphones and knowing global sales (which Samsung reports) and finally knowing global total phone shipments allows us to price the average non-smart phone.

Excluding the non-accused smartphones, the average revenue per smartphone in the US for Samsung in Q2 was $339. Adding in the estimated Galaxy SIII (4 million) at an average price of $470 gives a new "ASP" of $413.

This is where the analysis goes off the rails.

If we assume the global ASP is equal to the US ASP and knowing everything else about the mix of smartphones (54% of total) and revenues (as reported) yields a negative ASP for feature phones. Assuming the revenue figure is correct, either the number of smartphones is wrong (50.2 million) or the global ASP ($413) is wrong. The total phone number is also an analyst estimate which can be called into question.

There are many assumptions being multiplied here so I have very low confidence in the result, but we should ask "what do you need to believe" in order for the 50.2 million figure to be real. The following scenario for example has to be true:

1. Global Samsung phone shipments total 93 million units (which means 42.8 million feature phones sold as well as the 50.2 million smartphones).
2. 6 million smartphones shipped into the US, equivalent to about 50% increase y/y, roughly in-line with market growth
3. 4 million Galaxy SIII shipped (but not necessarily sold through) at an ASP of $450. This results in a US total smartphone ASP of $413
4. The global smartphone ASP is significantly lower (at 80% of the US or $330)
5. Global revenues of $18.14 billion is essentially all from phones
6. Global feature phones will have an average price of $36

If these conditions are met then it's possible that 50.2 million Samsung smartphones did ship.

The figure for feature phone ASP is difficult to accept as it would be lower than Nokia's. Also, the low global smartphone ASP is also difficult to accept but it might be possible to find a scenario with the above variables which makes the 50.2 million figure believable.

Unfortunately this is an unsatisfying analysis. I could propose another figure to answer the question in the title, based perhaps on more "believable" assumptions. But I have no means to defend my assumptions. The only option I have remaining is to caution that the total smartphones shipment estimate for Samsung remains precarious. The new evidence from the US trial places more doubt on it but it is not enough to disqualify it entirely.

---

Notes:

1. This cone of silence over 93% of the most popular mobile platform is somewhat puzzling in contrast to the perfect information we are used to having about iOS, BlackBerry and even Symbian. It seems that the more open a platform, the less is revealed about its market performance. This extends to the licensed Windows Phone whose shipments are only revealed through Nokia's reporting of Lumia shipments.
2. In the PC market, vendors also selectively publish their units. The norm is, however, not to. As a result IDC et. al. are the only sources for global market data. However, the PC market can be sampled more easily than the phone market. PC retail measurement methods are more reliable compared with all the points of sale for phones.
3. Note that Apple sells about 30% of iPhones in the US.
4. Apple's pricing world-wide is very consistent with pricing in the US which makes sense from a channel management point of view. One would not want to create incentives for cross-border arbitrage of phones.