asymco-admin

Sean Devine has a conversation with Horace Dediu of Asymco about open source software.

via 5by5 | Ruby on Rails Podcast #146: Horace Dediu of Asymco - Open Source Software.

This week Myke is joined by Horace Dediu. They discuss his work at Nokia and how that lead to starting his blog, Asymco. They also talk about the role and work of an analyst, his presentation series – Airshow – and graphs, naturally.

via 5by5 | CMD+Space #92: A New Lingua Franca, with Horace Dediu.

There are 7.1 billion people on Earth. Coincidentally there are also 7 billion mobile connections.  Those connections are held by 3.45 billion unique mobile subscribers. ((GSMA)) Unsurprisingly, the largest national mobile markets (by number of subscriptions) correspond closely to the most populous nations.

Considering smartphones, last year 1 billion smartphones were sold and the number of smartphones in use is about 2 billion ((There are also about 2 billion 3G/4G connections world-wide))

Given the rapid adoption of smartphones, it's also safe to assume that smartphone penetration will follow population distribution. In the US, where comScore data is published monthly, penetration is following a predictable logistic curve.

 

Assuming similar patterns world-wide we can forecast regional smartphone penetration.

This yields the following forecast for smartphone usage world-wide.

Three years ago the notion of smartphones becoming as ubiquitous as toothbrushes seemed far fetched but it's now gradually being accepted. The smartphone is well on its way to becoming the most successful technology in history.

So what?

What will the smartphone do for us? Does smartphone ubiquity change anything? Besides increasing ARPU slightly in a $1.2 trillion industry? Besides doubling the size of the device business? ((

))

Besides creating a $20 billion app industry? How do we assess the impact of having a connected computer in one's pocket? In everyone's pockets? With us all day. Every day.

Before we answer, we have to realize that the smartphone has been disruptive. Which means it has that special quality of not affecting the world in a predictable way. Its growth has been largely unforeseen and has affected the fortunes of the incumbents. As big and ubiquitous as it has become, the companies which were most intimately aware of it, its very own creators, were cursed by it.

The evidence is in this graph:

 

As the market exploded, the incumbent platforms all but disappeared. The innovators fell into the trap of treating the complex new product as a niche whose appeal was for selected "segments" described by the coarse data they were able to gather. Ubiquity itself was asymmetric to the basis of competition they established.

Today we can note how even the usurpers iOS and GMS Android are being swamped by the more modular alternative AOSP. Indeed, if we squint a bit we might notice that the OS/platform churn we see is part of a trend where the value is migrating up the stack from Components and Networks, through Devices and Operating Systems and on to Services and Experiences.

I call these the three mobile eras: (Firm names used are representative of a cohort and are not a complete list of participants.)

The thesis is based on Value Chain Evolution Theory which states that disruptions cause profits to evaporate from commoditized segments and condense in adjacent spaces, usually toward more modular, more convenient or lower cost innovations.

Note that the eras have specific time spans. If the pattern holds then we can overlay the time frame of smartphone penetration on these eras and even plot ecosystem entrants.

 

Disruption suggests that we are likely to see yet more shifting of fortunes. It means enabling new ecosystems to take advantage of the modularization of platforms. But how can it happen given the entrenched nature of the incumbent duopolists?

The clue lies in the first graph in this post: It suggests a geographic segmentation. The largest markets are large enough to sustain independent ecosystems. Note that China and Russia have already established alternative local service ecosystems. We may dismiss the possibility of localized Google and Facebooks in emerging markets but China and India are universes in and of themselves. But there are another 10 markets of more than 100 million users. That's another billion users among them.

But it's not just the availability of a vast pool of new users. The rapidity of growth of services like WhatsApp and SnapChat and Tencent shows that barriers to entry are low. We see increasing indigenous innovation in payments, logistics and business models in the newest markets. We see no shortage of talent and capital to start these businesses. We see increasing language and culture clustering in social media. We see political resistance to hegemony.

Focusing on the established, highly penetrated markets as an indicator of opportunity would be a mistake. Those markets foretell the rate of adoption but not the evolution of the value chain. Apple and Google will continue to anchor in those markets but there is no reason for the remaining two thirds of the world to follow. Indeed, there are incentives not to.

Incentives exist to localize services. Indeed, service businesses have always been local businesses. There are few global service brands. This has been true in retail, transportation, banking, hospitality, media and communications. Why should we believe that ecosystems can overcome the advantages of localization? Services are local because they are delivered by individuals and have high "touch" with the user.

Fundamentally, ecosystems are social systems and social systems have cultural and political dependencies. ((Post-modernism is a typological world view which  sees truth as socially constructed and a rejection of 'totality')) Technology does not homogenize unless it is centrally controlled. The control of mobile computing has already been abdicated by its creators.

We talk about the major triumphs and minor failures of the Veronica Mars campaign on Kickstarter and kick off a series on The Capitalist's Dilemma.

via 5by5 | The Critical Path #114: Veronicas Dilemma.

We talk about CarPlay, MWC2014, Microsoft and Nokia, "Tim Cook's outburst", per-user value, and re-evaluating how we analyze companies in tech.

via 5by5 | The Critical Path #113: Takt Time.

When trying to assess the success of an ecosystem, the primary measure is the size of the user base or the "audience" for the product. Companies like FaceBook and WhatsApp and Twitter are measured first and foremost on this metric. Companies like Google, Amazon, Apple and Microsoft are less so. When revenues are firmly attached to products the focus shifts to "follow-the-money" rather than "follow-the-users" metrics.

That's as it should be, but for the sake of understanding the competition between ecosystems, they should be compared on some similar basis. If the basis of competition in this day and age is ecosystems ((As Microsoft, Apple and Google seem to agree it is.)) how does one evaluate Facebook's vs. Microsoft's? Or, more poignantly, how does one compare WhatsApp's valuation with that of iMessage?

It's common to value a company which aggregates audiences at a multiple of that audience size. The implication being that each member of the audience returns a certain cash flow to the aggregator and the discounting of those flows is the net present value. Which is why, for example, a newspaper is valued in terms of its subscribers. As is a TV network and as might be a social network. Using this metric, a WhatsApp (i.e. free messaging) user is worth $40 and the average social media user may be worth around $100. ((The degree to which a user is actually worth the sum of all the subscriptions she may have made is a good topic for discussion.))

The implication is that users/subscribers/audience members are loyal and will stay with the programming for some time. There is also a second implication that businesses which are not measured by audience size don't have this loyal and recurring revenue base. The absence of an "audience" implies transience and impermanence and results in deep discounting of long-term viability.

Which is why ecosystems are the desired business construct for technology companies. They allow a more consistent and repeatable transaction model and offer predictability which is sorely lacking in the rapidly changing technology industries. It allows a company to define itself on its customers rather than on its products.

However, old habits die hard. Regardless of the exhortations of product companies, the notion of being valued on audiences is reserved for the service-oriented. Hence the outburst from Tim Cook at the latest shareholder's meeting. Reports have him saying that:

Apple now has 800 million iOS users and has handled over 16 trillion push notifications, with 40 billion new ones occurring every day. Apple also delivers several billion iMessages and FaceTime requests every day. Cook added that iPad now accounts for 78 percent of all tablets used throughout the enterprise.

Trillions of push notifications aside, the claim that Apple has 800 million iOS users seemed way out of line. This is because the number of iOS devices (excluding Apple TVs) shipped through to the end 2013 was exactly 774,336,000. One could easily imagine another 25 million or so since, giving us about 800 million sold, but I cannot see the number in use being the same, and even if they were, the number of users being the same.

The following graph shows the claim in context of shipments.

Note that last September Cook made the claim that 700 million iOS devices "by October", which did indeed happen. The 800 million claim in February is certainly believable in that context.

Determining the total number in use is a very speculative exercise. Assuming useful life of 9 quarters for iPhone and iPod and negligible iPad retirements yields about 560 million iOS devices in use. Obviously this can be very inaccurate as also there can be multiple users per device but the claim of 800 million users seems very hard to reconcile.

Perhaps he was mis-quoted or he mis-spoke but the real question isn't the exact number but rather the implication that Apple should be seen for its users rather than its quarterly shipments.

Them's fighting words, Mr. Cook.

[Update: Thanks for Philip Elmer-DeWitt's investigation, Apple clarified that Tim Cook referred to 800 million as the number of iOS devices sold not users of iOS. He was mis-quoted.]

The orthodox vs. the unorthodox: Tata, Tesla and Toyota. Why might an asymmetric competitor lose and a symmetric competitor win?

We begin with Tesla and Apple. We continue with aluminum vehicles and re-visit information asymmetry as Horace exploits it to buy a Mercedes on eBay.

We talk about car APIs (Aux input jack and ODBII) and much, much more.

A brief discussion considers the perils of endless line extension up and down the market, perhaps fueled by financialization.

This is a good one.

Asymcar 10: Asleep at the switch | Asymcar.

One of the axioms of hardware business is that prices fall over time. The consumer price index for personal computers and peripheral equipment from 1998 to 2014 is shown below:

The price index suggests that prices for computers should be 54% of 2007 levels. Charles Arthur illustrated this on a global basis using a separate set of data.

The data shows that the weighted average selling price (ASP) of a PC has fallen from $614.60 in the first quarter of 2010 to just $544.30 in the third quarter of 2013, the most recent date for which data is available.

The result is a chronically low margin for computer manufacturers. As Charles points out, the average profit for a PC is less than the price of a coffee at Starbucks. LG and Sony's exits from the market are symptomatic of the crisis. This is not a new phenomenon. There has been a long series of mergers, acquisitions and exits by incumbents.

Except of course for one incumbent.

The Mac, aged 30, seems to soldier on. Phil Schiller noted that of its original competitors, Apple is the only PC maker to still be in the market. One could imagine that this existence is marginal, or that the Mac managed to survive by giving up on unit volumes or pricing.

The following graph shows the revenue per unit for the Mac and its volumes.

The Mac not only continues to preserve pricing but it also seems to grow volumes at a time of overall market decline. Going back to the consumer price index, the Mac in December 2007 was priced, on average, at $1539. In December 2014, the price was $1322, with $40 deferred.

So rather than being 54% of December 2007 prices, the Mac is almost at 90%.

You'll note that along with the Mac I also illustrate the pricing of Apple's other devices. Even with the iPod, the pricing has remained stubbornly consistent for over 7 years. Whether the volumes are falling, rising rapidly or remaining flat, Apple's product pricing remains predictably consistent.

The point being that the violation of the axiomatic price erosion rule in PCs is not anomalous. It suggests something about the violation of the same rule in devices. It suggests that perhaps the company does not respond to the same market forces.

But we already know that. Jony Ive told us as much:

“We’re keenly aware that when we develop and make something and bring it to market that it really does speak to a set of values. And what preoccupies us is that sense of care, and what our products will not speak to is a schedule, what our products will not speak to is trying to respond to some corporate or competitive agenda. We’re very genuinely designing the best products that we can for people.”

The result is plain to see: Caring about the product means that it can be priced at a point which consumers care to pay.

Trouble is, judging by how Apple is valued, nobody believes that this is sustainable. Regardless of the evidence within Apple's own history, it's the exception, not the rule and the bet continues to be that they cannot continue to remain an exception.

Is Apple going to continue to repeal the laws that govern its industry, as it has for decades, or will it, as the market expects, abide by them?

Prior to Apple's earnings report I read at least one article suggesting that the most important indicator to watch was Apple's margin. I suppose this was due to a recent decline in margins from a peak gross margin of 47.4% in Q1 2012 to 36.7%.

As the graph below shows, margins began to recover by Q3 2013 and are nearly on par with year-ago levels.

The guidance for the present quarter is a gross margin between 37% and 38%. This would imply a flat q/q GM line (blue line above.)

This is not quite catastrophic.

To better understand margins, it helps to compare them with other companies. When Apple reached that peak of near 50% gross margin I noted that such a level was higher than Microsoft's and Google's. The irony being that Apple was nominally an (implied) low-margin hardware company while Microsoft was an (implied) high-margin software company and Google was an (implied) high-margin internet services company.

Here is the picture with the last two years added:

In the last two years Apple's margins have declined but so have Google's. And although Microsoft's fell, they fell less than Apple's and thus overtook it to remain slightly more profitable (32.5% vs. 30.3%) into Q4.

I also added Samsung Electronics' operating margin. ((As a large conglomerate, Samsung will have a lower blended average. Separating the Telecom group's operating profit shows a 16.14% divisional operating margin.))

So although Apple stabilized, it did so at an extraordinarily high level, well out of band for a hardware company. Thus the question remains: Are Apple's margins a fluke? Is the recent rise of margins to levels far above those of a comparable hardware company and into the realm of software and services sustainable?

The simple answer is to assume that the iPhone and iPad, which make the bulk of revenues and which explain much of the rise, enjoyed a temporary buzz which allowed them to be priced at a premium but as that buzz wore off, the pricing and margins would decline. This is common with fads and fashions.

But the situation is far more nuanced. If the margins were due to mere novelty, the effect should not have persisted over so many years. Pricing data shows that the iPhone remains exceptionally valuable and that the iPad, excluding line extensions, has also held the line on pricing.

The missing piece in the puzzle is that the iPhone and iPad and iPod touch and Apple TV are pieces of an inter-dependent network of assets. The iTunes stores, Apple's services and Apple Retail are easily ignored because they are not "profit centers" but I believe they are a key component to the sustainment of these margins.

In other words, the iPhone and iPad enjoy margins not just because they are new and well designed. But also because they form part of a ecosystem. The key word is 'system'. The user buys into a system not just what the device embodies.

This is why I've always said that the iOS product businesses should be thought of as "hardware as a service". The inspiration comes from the notion of "software as a service" which posits the idea that software purchases should be thought of as services and thus recurring revenues.

iOS devices are part of a service matrix: the user experience in the stores, the after-sales support (including the easily ignored Applecare), the broadband supplied through operator service (which is part of a trillion dollars of value), the apps and content that can be purchased (which add up to a mountain of money), the network effects of iMessage and iCloud and now iRadio and Match and Siri.

These services attachments are well understood as great business models and underpin some of the most celebrated businesses to emerge in the last decade including Google, Amazon, Facebook, Box, Dropbox, Windows Live, etc.

It's curious therefore that whereas these other businesses are highly valued as services, in the case when the service is a component within a larger system, the value is missed. In an integrated context the connection between services and high margins is not understood. The reason may be the absence of robust systems analysis and a preference for the facile explanation.

Horace Dediu and Guy English discuss the future of TV, how we consume media, and the fate of console-style devices. From Apple to Nintendo and beyond, is there one box to rule them all, or can there be?

Interview segment: Cinematographer Barry Braverman talks about working with director Wes Anderson (uncut version to be posted as Screen Time #54)

via 1: Bicycle for the Mind.

• Apple TV Gains its Own Shopping Category | AppleInsider
• Horace wrote one of my favorite Asymco articles in 2011 that applies here: Hiding in Plain Sight
• AirPlay Configuration Files Hint at Next-Gen Apple TV Hardware |AppleInsider
• Mobile Gamers Aren't Playing NIntendo-Style Games | ConsulGamer
• Barry Braverman's website, and his IMDb credits.
• Bicycle Thieves, aka The Bicycle Thief aka Ladri di bicciclette | Wikipedia
• Bottle Rocket (and Murita Cycles) | The Criterion Collection
• Umberto D (click through for restored trailer) | The Criterion Collection

Horace and Moisés discuss the excitement surrounding Microsoft's hire of new CEO/poet Satya Nadella, Apple's recent numbers announcement, and yet more on Google (and their sale of Motorola to Lenovo).

via 5by5 | The Critical Path #111: Man of Arts and Letters and Science.