stratechery
stratechery

Podcast: Exponent 009 – The Societal Perspective

On the newest episode of Exponent, the podcast I co-host with James Allworth:

In this episode we first defend surge pricing before continuing our debate from last week: is society better off when corporations try to innovate, or would we be better off if we left the innovation to startups? Your answer to this question may change your opinion of Silicon Valley.

Note: In the podcast I talk about owning Microsoft stock; to be clear that was while I was an employee. As a matter of policy I do not currently hold any individual stocks. Also, this podcast was recorded before Satya Nadella’s recent strategy memo, which I analyzed in today’s Daily Update (members only)

Show Link

Feed | iTunes | Twitter | Feedback

Sponsor: MailChimp

Freddie_OG

Send Better Email. More than 6 million people and businesses (including Stratechery!) use MailChimp to design and send email marketing campaigns.

Join them today.

My thanks to Mailchimp for sponsoring Stratechery this week.

Smartphone Truths and Samsung’s Inevitable Decline

For me, anyway, the most surprising thing about Samsung’s disappointing earnings was just how surprised many folks seemed to be. The smartphone market is a massive one, but also rather predictable if you keep just a few key things in mind:

  • Everyone will own a smartphone – I don’t think this is controversial, but it’s important, as there are a few implications of this fact that are perhaps non-obvious.

  • The majority of buyers will prioritize price – The implication of a phone being a need and not a want is massive downward pressure on the average selling price for two reasons:

    • Low income buyers who might normally not buy consumer electronics or other computing devices will be a part of the phone market, and will buy low-priced models by necessity
    • Higher income buyers who are uninterested in other consumer electronics or other computing devices will be a part of the market, and will buy the low-priced models by choice

    The net effect is that the average selling price for a phone will be low and always decreasing, while the high-end market will be relatively small in percentage terms.

  • Absolute numbers matter more than percentages – While it’s natural to talk about market size as a percentage, the absolute size is just as important. In the case of Apple, for example, the fact they “only” had 15.5% percent of the market in 2013 is less important for understanding the iPhone’s viability than is the fact they sold 153.4 million iPhones. That is more than enough to support the iOS ecosystem, percentages be damned.

  • There will always be a high end segment – The very reason why everyone will buy a phone (always with you, access to information, communication) are the same reasons there will always be a segment of the population willing to pay for a superior product. The analogy to cars is perhaps overdone, but for good reason: it makes a lot of sense. Like cars, phones are about appearance, performance, and experience; both are status symbols; and (in most parts of the world) both are necessities.

  • The high end isn’t that expensive on an absolute basis – Where the car analogy breaks down is absolute prices. The cheapest Mercedes-Benz you can buy (in the U.S.) is a surprisingly accessible $29,900. That, though, is 46x as expensive as an iPhone 5S. Sure, an iPhone 5S is a bit more than 3x as expensive as a Moto G, but the absolute price difference is only about $500; a car 1/3 the price of that Mercedes would have an absolute price difference of $20,000.

  • Low end quality is improving rapidly – That Moto G is a very nice phone that absolutely does the job for most people. It’s also not that big a deal, particularly in Asia where there are even cheaper and more capable phones available based on a SoC from MediaTek. Moreover, the entire supply chain continues to improve and bring down prices on every part of a smartphone, improving the quality of even the most inexpensive products.

  • Fleshed-out App Stores are table stakes – By fleshed-out app stores, I mean the iOS App Store and Android’s Play Store, full stop. It’s nice that Windows Phone and the Amazon Fire app stores are getting some big names, but the problem with an 80/20 approach (in this case, shooting for 20% of apps that satisfy 80% of needs) is that everyone differs on the remaining 20%, and it’s usually that 20% that is the most important for any one user. Of course, some users don’t really care, but those are likely super low-end customers anyway, which aren’t great for margins or your ecosystem.

  • Carriers matter, at least for the high end – Many customers, particularly in developed markets, are loyal to their carriers and only choose phones which are available on their preferred network. On the flipside, markets in which people move freely between carriers (or use dual-SIMs) are usually lower-income markets with smaller high end segments.

  • Screen size matters – The one physical characteristic that seems to impact phone selection is screen size. While the large screen phones are a relatively low percentage of the total phone market, they are a much higher percentage of the high end.

  • Software Matters – For years analysts treated all computers the same, regardless of operating system, and too many do the same thing for phones. I personally find this absolutely baffling; you cannot do any serious sort of analysis about Apple specifically without appreciating how they use software to differentiate their hardware. The fact is that many people buy iPhones (and Macs) because of the operating system that they run; moreover, that operating system only runs on products made by Apple. Not grokking this fact is at the root of almost all of the Apple-is-doomed narrative (which, by the way, is hardly new).

    Software-based differentiation extends to apps. While a fully-fleshed out app store is table-stakes, for the high end buyer app quality matters as well, and here iOS remains far ahead of Android. I suspect this is for three reasons:

    1. The App Store still monetizes better, especially in non-game categories
    2. iOS is easier to develop for due to decreased fragmentation
    3. Most developers and designers with the aptitude to create great apps are more likely to use iOS personally

    None of these factors are likely to go away, even as Android catches up with game-based in-app purchases and as iOS increases in screen size complexity.

It is this final point that makes the Samsung news so unsurprising. Samsung had built up a healthy high-end business by:

  • Being available on nearly every carrier
  • Pioneering the large-screen segment
  • Producing hardware that was meaningfully superior to low-end offerings

All three of these factors either have or are in the process of disappearing:

  • After a two-year lull, Apple has greatly expanded iPhone availability worldwide
  • As noted above, the gap between low and high-end hardware is disappearing
  • Multiple manufacturers have moved into the large-screen segment, with the iPhone coming soon

In China Samsung has another problem: their brand and distribution channel, which they have spent billions building, is no match for Xiaomi’s star power which lets the startup sell phones at cost without any additional marketing or channel expenses. It’s not helpful to (rightfully) say this issue is primarily limited to China (I’m more skeptical of Xiaomi’s prospects elsewhere, but bullish on Lenovo) because the Chinese market is the largest market in the world.

Ultimately, though, Samsung’s fundamental problem is that they have no software-based differentiation, which means in the long run all they can do is compete on price. Perhaps they should ask HP or Dell how that goes.

In fact, it turns out that smartphones really are just like PCs: it’s the hardware maker with its own operating system that is dominating profits, while everyone else eats themselves alive to the benefit of their software master.

Previous articles on Samsung’s troubles:

Happy Independence Day Mr. Glass

Blessed with the sort of love him or hate him reputation reserved for the truly popular, Bill Simmons has received a lot of criticism from NBA fans for his propensity to act as the Body Language Doctor: he will make grand pronouncements about players or teams based on nothing more than a player or coach’s demeanor on the floor or in a press conference.

Still, I can sympathize with Simmons: it’s easy to give in to a similar sort of temptation when you read things like this New York Times article about Ira Glass and This American Life leaving Public Radio International:

On July 1, “This American Life” became independent, leaving its distributor of 17 years, Public Radio International, or PRI. That change is partly technical. The program is no longer delivered to local stations through public radio’s satellite system, but instead over the Internet through the online platform PRX, the Public Radio Exchange.

But the big impact is financial. Gone are a distributor’s financial guarantees, which in the case of “This American Life,” reached seven figures. Instead, Mr. Glass will now be responsible for the show’s marketing and distribution, as well as for finding corporate sponsors. It’s the equivalent of Radiohead’s releasing its own album “In Rainbows,” or Louis C. K.’s selling his own stand-up special — except all the time, for every show. It’s the kind of move that can signal radical changes in the public radio firmament, with National Public Radio and other distributors wondering who, if anyone, may follow suit, and whether Mr. Glass will return if he fails.

“You take on the risk if you have to do the marketing,” said Laura Walker, president and executive chief officer of New York Public Radio, which operates WNYC. “I don’t think it’s a slam-dunk way of making money. You’ve got to put in a lot of effort and do the work yourself.”

Set aside the implications for Glass and This American Life for just a moment: what is striking about the article, and this section in particular, is that there is zero discussion about upside. The “big” financial impact is the foregoing of financial guarantees, and questions are raised about what happens if Glass fails – but not if he succeeds. There is concern that Glass’s move towards independence is not a “slam-dunk way of making money.”

To be fair, this is only one article, but the reason the body language doctor angle is so tantalizing is that this approach seems very representative of traditional journalism’s general discomfort with the Internet. When your world is collapsing it’s awfully easy to see only the downside, and to wish that things like disruption did not exist.

So let me provide a counter-narrative, and re-write the lede to this story:

On July 1, just days before the country he chronicles marks Independence Day, Ira Glass of This American Life celebrated his own independence, leaving his distributor of 17 years, Public Radio International, or PRI. That change is partly technical. The program is no longer delivered to local stations through public radio’s satellite system, but instead over the Internet through the online platform PRX, the Public Radio Exchange.

Of course This American Life is no stranger to Internet distribution; while 2.1 million people listen to the show live on the radio, another million download the podcast, making it the most popular show on Apple’s iTunes. In fact, it was Mr. Glass’s ability to connect directly with his show’s listeners that made an intermediary like PRI redundant.

To be sure, Mr. Glass is taking a risk by abandoning a distributor’s financial guarantees, which in the case of “This American Life,” reached seven figures. Instead, Mr. Glass will now be responsible for the show’s marketing and distribution, as well as for finding corporate sponsors. The upside, however, is enormous. The cost of a financial guarantee is limited upside – that is why distributors take on that risk – but by taking control of distribution Mr. Glass is reserving that upside for himself.

“The entity with the most to lose in this move is not This American Life,” said Ben Thompson, who has written frequently about the impact of the Internet on journalism at his blog Stratechery. “Rather, once other radio personalities realize that the Internet has made distributors redundant a lot more people are going to question why they don’t take control of their own destinies.”

The clear winners, though, are consumers: Mr. Glass has marked the occasion by releasing a new podcast called Serial, and it will be available to everyone worldwide, no distribution deal needed.

Happy Independence Day to Ira Glass.

And, dear readers, Happy Independence Day to you as well, both you in the U.S. actually celebrating the holiday, and also everyone around the world who I can reach with ease, no distributor needed.

Additional Thoughts on iWatch and Android Wear

A few points of follow-up on last week’s piece Android Where?:

So What About iWatch?

I only mentioned the iWatch tangentially in last week’s piece, which is just as well, for it gives me an opportunity to link favorably to this piece on Techpinions by industry veteran Tim Bajarin about the (alleged) iWatch:

I actually think the ID aspect of any wearable Apple brings out is probably central to its future functionality. This is speculative on my part but, after using the Disney band for seven days and seeing its incredible functionality, Apple has to be crazy not to make this part of any of their wearables. The ramifications for Apple’s future with this one ID implementation alone could make it a huge hit. Imagine going into a Starbucks and just touching your iWatch or iBand to the terminal, entering a PIN number and it is charged to your Apple account. Or to enter your house, you just touch the Apple wearable and enter a PIN number and your are in. Yes, you can do this with an iPhone now but that means taking it out of your pocket or purse and it is only single authentication at present. In a wearable, it is much easier to use for entering the home and for all types of interactions and transactions. Its convenience factor would be very compelling. I believe something like this would be very “sticky” and keep users of these tied closer to Apple’s ecosystem…

I believe when and if Apple does launch an iWatch or iBand or what ever form of wearables they bring to market, they will initially lead with the health and home automation apps first and over time add the ID features. As you can imagine, using an iWatch or iBand for ID that can be used to do transactions, lock doors and even handle proximity functions could be controversial without them first convincing people they can trust them even more than they already do today. That could take some time.

I agree almost completely with his piece, and encourage you to read the whole thing. In fact, it’s exactly what I was driving at in Android Where? when I said:

I question, though, if Now will turn out to be as meaningful to most people as Google thinks it will be. In other words, how many people actually want a personal digital assistant? There is an alternative view of computers in which they are more akin to a tool, something you pick up and use to do a job, and then set down when you are done with it. To be sure, that tool is incredibly powerful and capable of doing a great number of jobs, but it still operates in service of something outside of computers…A digital assistant, on the other hand, is simply a more efficient way of interacting with your computer – or your computer interacting with you – and I question how much that vision that will ultimately resonate with people.

The potential problem with Android Wear I see is its focus on being a conduit for your phone – a more efficient distraction from the real world, as it were. I suspect the iWatch will go in the opposite direction: it will be a means of pulling the real world into your phone (think sensors) as well as a means of projecting your phone out (think payments, proximity, etc.)

Interestingly, neither of these use cases necessarily even need a screen. And, if that’s not a requirement, the range of possible objects increases exponentially. It could be a ring, or a bracelet, or a clip, and so on – or all of the above. It would also be great for battery life, and for price (touch screens are the most expensive component in phones and tablets). Remember the Wall Street Journal article from a week-and-a-half ago:

Apple is planning multiple versions of a smartwatch—dubbed the iWatch in the media—later this year, according to people familiar with the matter.

One more thing: a portfolio of wearables certainly would explain hiring someone like former Yves Saint Laurent CEO Paul Deneve as head of special projects. Regardless, I think it’s very safe to say whatever Apple releases will look nothing like the devices shown by Google last week.

Android Wear/TV Can’t be Customized

One interesting detail about Android Wear (and Android TV) is that Google is prohibiting OEM’s from customizing the look-and-feel. From ArsTechnica:

One thing about both of them sticks out: their software behaves pretty much the same way no matter which device you have. There are small differences that Google has outlined here, but interacting with each watch is exactly the same, and digging down into the settings shows that they’re both running the exact same Android versions and build numbers. This would be unusual for Android phones or tablets, which generally come with OEM-controlled UI skins, hardware and software flourishes, and pre-installed apps.

Talking with Google engineering director David Burke confirmed that all of the new Android initiatives announced at the keynote this week—Android Wear, Android Auto, and Android TV—will have user interfaces and underlying software that is controlled by Google, not by the OEMs.

“The UI is more part of the product in this case,” Burke said to Ars of Android TV in particular. “We want to just have a very consistent user experience, so if you have one TV in one room and another TV in another room and they both say Android TV, we want them to work the same and look the same… The device manufacturers can brand it, and they might have services that they want to include with it, but otherwise it should be the same.”

So here is my question: if “the UI is more part of the product” when it comes to Android Wear devices, does that mean the UI is not part of the product when it comes to Android phones?

Obviously that is not the case in a strict sense – UI is incredibly important in a phone – but looked at broadly Burke’s statement does have a grain of truth. The fact Google does not lock down the UI for Android on phones is not because they don’t think it is important, but rather that the context in which Android was launched was a far different one. As I wrote after last year’s Google I/O, Android was originally developed not as an iPhone alternative, but as a Windows Mobile alternative. Google wanted to ensure that Microsoft did not dominate mobile in the same way they dominated the desktop.

To that end, they needed to have a “better” offering that Microsoft, where by better I mean an offering that was more appealing to OEMs (remember, back then Microsoft was using their PC business model for mobile – OEMs were their customers). Part of this was “free,” but part of it also was the promise that OEM’s could create unique experiences. Microsoft had been seeking to lock down PC Windows for years at this point, and had taken a much stricter approach to Windows Mobile.

What’s interesting, of course, is how that last sentence perfectly describes Google’s approach to Android and Android Wear. Increasingly locked down, and, now that they are dominant, their new platform can’t be customized at all.

(As an aside, while I have said this multiple times, it’s worth repeating: Android was a strategic masterstroke. It’s not just that Google now controls the platform on the vast majority of the user’s phones and all the power and data that come with that, but also that mobile’s emphasis on apps over the browser would have severely magnified Microsoft’s chokehold had they been as dominant as Google feared.)

Parenthood iPhone Ad

Back to the iWatch, there is a great new iPhone ad by Apple. From Recode:

Today is the seventh anniversary of the very first iPhone release and Apple is celebrating it with a new ad touting the device’s role in the Internet of Things and its ability to support parenting and foster family interaction. Showcased in the spot are all manner of family-friendly peripherals: the WiThings baby monitor, the Belkin WeMo plug, the ProScope Micro Mobile microscope, the Tractive GPS dog collar, the Kinsa thermometer and Parrot’s Flower Power wireless sensor for plants, as well as a bunch of related apps.

A few weeks back in my Daily Update (members only) I wrote that Apple’s new fitness ad, while being nominally about the iPhone, probably ought to be viewed as the beginning of the iWatch marketing campaign:

As countless commentators have noted, one of the central challenges of the alleged iWatch is that it’s not super clear how big of a need there is among the general population. What, though, is advertising? In many respects the best sort of ads make you aware of a need you didn’t realize you had. This sort of effect, though, is not achieved with a one-off spot or campaign. Rather, it’s a long slog that only sees results over time.

To put it more bluntly, it’s very possible that we just saw the first iWatch commercial. Oh sure, there is no iWatch to be seen, and there are benefits depicted in the commercial that accrue to the iPhone today, but I wouldn’t be surprised if this is Apple starting to set the table amongst the broader consumer market.

To my mind this new ad confirms that thesis. Apple is working to set the expectation that your iPhone is better with additional peripherals, and golly, wouldn’t it be something if Apple had a peripheral of their own to sell you?

Beyond that, as a parent I really did enjoy this ad, and felt it hit the right notes. That said, they did manage to skip over the “use-an-iPhone-playing-YouTube as a sedative” angle. :)

Android Where?

There is no questioning the scope of Google’s ambitions. Consider the fact that yesterday’s two-and-a-half hour keynote was shorter than the 2013 version, and that the two keynotes had almost zero overlap!

While last year Google spent a lot of time both enhancing the Android development environment and Google+, this year was about extending Android to a whole host of new devices, including watches, cars, and TVs. However, just because Android has been an unqualified success on phones – and deservedly so – I’m a bit dubious as to how well Google’s vision will translate to the next generation of devices, in particular Android Wear.

Note: I reviewed all of Google’s announcement in today’s Daily Update (members-only)

Watches Are Not Phones

Beyond the rather obvious fact that one is on your wrist and the other is in your pocket, the primary difference between a watch and your phone is that you buy the former by choice and the latter by necessity. This has a significant effect on not just the size of the addressable market, but also the type of buyer in that market.

Because a phone is a necessity – everyone is going to buy one – there is going to be a very large segment of the market whose purchase decision is driven primarily by price. In turn, this means the minimum acceptable product with the lowest price is going to garner the largest market share. Indeed that is exactly what has happened with phones: while the iPhone is actually increasing its dominance of the high end, Android has covered the entire rest of the market. And, as we learned yesterday, that market is now >1 billion in size (and the purpose of Android One is to capture the huge swathes who cannot or will not pay more than $100).

To put it another way, in the case of a phone, the only decision to make is which phone to buy. When it comes to a watch or other wearable, though, there is another question that has to be answered first: do I even want to buy one at all.

This additional question will make wearables a much more difficult market for Android to crack. Instead of simply making available an item the customer needs to buy at a low price, the wearable maker must first demonstrate a reason to buy into the category in the first place, and only then does price come into play. Moreover, any item that is purchased because you want it is a sort of luxury good, at least relative to something you buy because you need it, which by extension means the percentage of buyers who will prefer a premium option will be much greater. This is the primary reason why Android’s tablet share is much less than its phone market share, and its usage share lower still.

New Devices Need New Interfaces

In many respects Google’s push to take their dominant platform and port it to what’s next is reminiscent of Microsoft’s previous attempts to take their dominant platform – Windows – and port it to phones. Which is why the Pocket PC 2000 (a.k.a. Windows Mobile v1) looked like this:

A screenshot from Pocket PC 2000 (Windows Mobile 1.0)

A screenshot from Pocket PC 2000 (Windows Mobile 1.0)

We laugh now at the idea of shrinking a PC UI to a phone, but I got the same sense from the Android Wear UI:

Notifications on Android Wear

Notifications on Android Wear

It’s hard to see the notification UI clearly, but it basically looks the exact same as a phone notification, just shrunk down. And, just like the phone, you swipe to dismiss. Sure, it’s very cool how the notification then disappears from the phone, but at the end of the day it’s hard to escape the feeling that you’re simply wearing a second phone on your wrist. Is saving the effort of reaching into your pocket really worth the cost, not just in dollars, but also in comfort and fashion? I’m skeptical, and it’s hard to see a device like Android Wear being anything more than a niche product, just like Windows Mobile.

Is Now the Future?

Notifications are nice, but probably the biggest selling point of Android Wear is as a conduit for Android Now. It’s Now that to my mind really speaks to Google’s ambition: having mastered the world’s information, Google is now focusing on the individual. Now seeks to be your digital personal assistant, anticipating your needs and providing you the information you need when you need it (and, of course, building one heck of an advertising profile). And, when it comes to just-in-time information, the presentation of which is initiated by your device, a watch is in fact a rather compelling addition to a smartphone. After all, as long as your phone is in your pocket you may not see the reminder about your upcoming meeting, or that you ought to leave a few minutes early because of bad traffic. By virtue of being in a more accessible place on your body a watch may be superior for this sort of passive interaction.

I question, though, if Now will turn out to be as meaningful to most people as Google thinks it will be. In other words, how many people actually want a personal digital assistant? There is an alternative view of computers in which they are more akin to a tool, something you pick up and use to do a job, and then set down when you are done with it. To be sure, that tool is incredibly powerful and capable of doing a great number of jobs, but it still operates in service of something outside of computers.

Consider your Android phone: it’s a way to connect with friends and family, a way to entertain yourself, a way to capture memories, and a way to learn. It’s meaningful because it lets you do what you want to do better. A digital assistant, on the other hand, is simply a more efficient way of interacting with your computer – or your computer interacting with you – and I question how much that vision that will ultimately resonate with people.


During yesterday’s keynote Benedict Evans posted this tweet:

I think there is more than a grain of truth here when it comes to Google, but I do feel Apple thinks about computing differently. Recall Steve Jobs’ famous characterization of the Mac – a bicycle for the mind. While a bicycle makes a human more efficient, it still depends on that human for locomotion and direction. I think we see that today with Siri. Siri does not try to anticipate your needs or tell you what to do, instead she responds to questions delivered by humans, helping them to find the answers far more efficiently than they could on their own.

The true test, of course, will come when we see Apple’s wearable. All of the criticisms I have levied against Android Wear potentially apply to Apple’s offering, particular the lack of a reason-to-buy. More fundamentally, though, the philosophical question I have raised is at the root of which company you think is more likely to own the future. Google already has the best computer, and their lead is only increasing; Apple, though, has always made the best bicycles.