Services

In the earnings Q&A Stephen Elop used the word 'Platform' once, 'Operating System' three times and 'Ecosystem' 15 times.

Though they seem interchangeable, the three phrases have significant differences in meaning. Briefly put, a Platform is a business model concept,  an OS is a technology concept and an Ecosystem is a marketing concept.

Platform

A platform, as the New York Times recently pointed out, can take many forms. The Barbie doll is a platform as is the US Interstate Highway system. These are products or infrastructure that allow other products to be 'built on top' and can thus charge rents from them. When building platforms, a company needs to think of how to make their work foundational so it can support other structures. Other examples of platforms are browsers, the Apple 30 pin dock connector, Game Consoles and Facebook.

In terms of innovation theory, a platform is a signal to industry that the company is willing to allow its innovations to be extended and thus share in the creativity of adopters. There is however a down side: platform adopters may also create so much value that users will shift loyalties to the layers above the platform thus restricting the freedom of the platform owner to modify or modernize the underlying platform.

This loss of control over the roadmap often forces platform owners to institute draconian limits to the extensibility of the platform (e.g. the iPhone developer restrictions.) In other cases, platform owners have acquired and integrated various core apps or services so that the platform and core apps can be upgraded in sync. This was the impetus to the development of Microsoft Office. Office was motivated by a need to prevent Lotus from slowing the upgrade path for Windows not, as it's assumed, to create a new revenue source. However, again, the effect is the destruction of the ability to innovate on that particular area (no new Word Processor or Spreadsheet programs for 20 years).

It takes a great deal of work to platformize a product. I would guess 10x more than a similar product which does not allow collaborative innovation. Platforms are difficult to build, require delicate balancing acts to maintain but can be enormously valuable due to network effects.

Operating System

In contrast an Operating System has a very clear technical definition. It has been extended to include many applications and services, but it's basically the control system of a computer. An OS may or may not be a platform. For example embedded operating systems typically do not have APIs to support applications. Some basic mobile operating systems support applications through a sandbox like JaveME (in this case the platform is Java) and are therefore also not platforms.

A company licensing an OS may not accept the platform with the license or may not consider it valuable. It's even possible to license an OS and consider the platform value to be a liability to your own strategy (arguably HP fits in this scenario as regards to Windows. HP had its own operating systems and tried to build alternative platforms in spite of Windows).

An example of an OS with minimal platform value may be WebOS which has a very limited API for native applications.

Ecosystem

The word ecosystem as it applies to technology was coined by Palm to describe the app vendors that built apps and the accessories for its Palm Pilot. This is not a common word as it's largely equivalent to Platform. The subtle difference is that it's used in a marketing context to signal platform value. Having a large ecosystem implies the acceptance of a platform by a large base of developers.

For example a platform may not have a viable ecosystem if there are few developers and/or little economic value 'on top' of the platform. Symbian is often cited as being a decent OS, but a poor platform with a mediocre ecosystem.

So these three phrases are overlapping but have a different tone. Although perhaps unintentional, Elop's emphasis on Ecosystem vs. Platform or OS, might signal that what's missing from Nokia's strategy is the motivation or encouragement of developers to participate. My prediction of WP7 licensing may not fit well with this deficiency but perhaps Microsoft's ecosystem building potential is still greater than Nokia's.

I should also point out that Nokia's platform strategy has also included an intermediate layer called Qt that enables cross-platform ecosystems, thus making the OS irrelevant. Technically this may have merit, but it's a long haul to getting an ecosystem for it.

60.7 million people in the U.S. owned smartphones during the three months ending in October, up 14 percent from the preceding three month period, representing 1 out of every 4 mobile subscribers.[...] Despite losing share to Android, most smartphone platforms continue to gain subscribers as the smartphone market overall continues to grow.

via comScore Reports October 2010 U.S. Mobile Subscriber Market Share - comScore, Inc.

Separately, Nielsen reports on the same market:

29.7 percent of U.S. mobile subscribers now own smartphones that run full operating systems.

The most popular smartphones are the Apple iPhone and RIM Blackberry, which are caught in a statistical dead heat with 27 percent of smartphone market share in the U.S. Twenty-two percent of smartphone owners have devices with the Android operating system.

via U.S. Smartphone Battle Heats Up: Which is the “Most Desired” Operating System? | Nielsen Wire.

I've been tracking the Nielsen data and using it as a basis for forecasting the penetration of smartphones. While most coverage of this data focuses on splits between current OS platforms, I will look at competition with non-consumption and the impact on operator services.

The chart on the left shows the historic penetration of smartphones since mid-2009 based on Nielsen's reports. The chart on the right is my forecast. This is an update from the last quarter's chart here. My forecast was fairly accurate with respect to platform shares in the last quarter so I've had to make only small adjustments and extended the forecast one quarter further out. [1]

If my forecast is correct, smartphones will make up 50% of all phones in use by the end of 2011. That's up from nearly 30% today and 18% a year ago.

One in three of all phones in use will be either iOS or Android. That adds up to about 80 million users; up from 35 million today. Overall, 120 million Americans will be using smartphones.

Now consider the consequences for how services are delivered by operators. The most popular product operators will be selling will be unlimited data plans. One year hence, by 2012, unlimited data plans may become the largest service revenue generator for mobile operators in the US. Pricing and competition will inevitably be in a state of extreme flux. I leave the consequences to the imagination of the reader.

---

[1] Note that the charts show share of phones in use, not phones sold in a particular point in time. In other words, this is measuring platform consumption, not units sold.

Google has created a partial list of phones that run Google services.

Google Phone Gallery.

I notice with interest that it's called a Google Phone Gallery not an Android Phone Gallery.

What's more, there is a distinction of All Phones and Phones "with Google". Of course that implies that we can get a list of Google Phones "without Google" (or maybe more accurately, "with partial Google").

To clarify, according to Google there are three categories of Android Phones:

• Google Phones "with Google" have been optimized for use of Google Mobile Services, providing easy access to Search, Voice Search, Google Talk, Google Maps, Gmail, Sync, YouTube and Android Market (where available).
• Google Phones "without Google" list includes Android phones that have the Android market, Search, Maps and Gmail. You can get this list by subtracting the Google Phones "with Google" from all the Google Phones.
• Non-Google phones. Android phones that have no Android market or any Google services. You can get this list by taking a list of all Android phones (e.g. from pdadb.net) and subtracting all Google Phones.

I would also add a fourth category of Google phones:

• Non-Android phones that serve up Google services just fine.

Since the list is not complete, I won't tally it here, but it suggests a large number of Non-Google phones is possible.

[I drew the following Venn diagram to help explain the situation. The shaded areas are those which generate revenue for Google. The interesting struggle (marked with arrows) is between the non-Android Phones with some Google and the Android Phones without any Google.]

Now the question that needs to be asked is: How many of the Non-Google Android phones are built to compete with Google (e.g. on Search, Voice, Maps, Media and App sales)? And a second question: How are the Non-Android phones "with some Google" competing with the Non-Google Android phones.

In other words, from Google's point of view: how successful are companies other than Google at delivering Google services in spite of competition from companies with Android phones delivering services that compete with Google?

The classic example being: how effective are iPhones with Google services at competing with Verizon Droid Android phones that deliver Microsoft services? In this scenario Apple is delivering value to Google shareholders while Verizon (using Android) is taking it away.

Is this trade-off a net positive?

In other words, is Google's income from the iPhone offsetting its losses from Android?

Despite being based on Android, a Facebook phone would be competition for Google as much as it is for Apple. Google benefits from ad revenue tied to their search and other services which would likely be supplanted by Facebook services on a Facebook-based phone.

Facebook to Enter Mobile Phone Market in 2011 - Mac Rumors. I wonder if Facebook devices, or Verizon Bing devices will be counted by Google as Android activations.

See also: asymco | Is a Facebook phone destined to be a Vanity smartphone?

France Telecom-Orange CEO Stephane Richard has invited the heads of Vodafone, Telefonica and Deutsche Telekom (T-Mobile) to Paris on 8 October to discuss the development of a common operating system for mobile devices. He told Le Figaro that mobile operating systems were the Trojan horse used by Google and Apple to establish relationships with mobile customers. The four operators have nearly 1 billion customers combined and the capacity to influce industry. The initiative could take several forms including a joint venture or a common apps development unit. He added that the operators aim to retake the reigns of innovation, rather than be followers.

In the interview Richard also stated that carriers have decided not to advertise the iPhone any more.

Phone carriers want to team up against Apple and Google. - HardMac Forum

This attempt at operator cooperation in software has been tried before, but it predictably failed (does anyone remember Savaje?) Symbian itself was an attempt at mobile phone vendor cooperation on systems software. There are various consortiums for "open mobile OS" based on Linux (e.g. LiMO which started as a Japanese effort). Even the Open Software Alliance is a rubber stamp body rather than an attempt at development.

Although one can safely dismiss the potential of this effort on the basis of a lack of competence, the arrogance reminds one of the rigidity in response to the disruption under way. The fact that high-speed data networks allow operator disintermediation cannot be changed by "innovation" on systems software.

But the odd thing is why even bother. The solution to taking control over the user experience and services platform is straight-forward: each operator could fork a version of Android as their own and hire a team to integrate white label services. It's much more straight-forward than coordinating a joint effort. China Mobile have already done their own non-Google Android.

As previously described in an earlier piece on the transistor, there is a theory that underlies much of what this web diary is about. Telecommunications is undergoing wrenching change. And not for the first time. The history of the industry is filled with disruption and thus with drama that leads to fine storytelling.

One such great story is that of Western Union and the telegraph.

Western Union was formed as a company in 1851 and ten years later had completed the first transcontinental telegraph line. Following on that breakthrough, it introduced the first digital real-time data service, the stock ticker, in 1866. Business boomed and by 1870 it was the world's most powerful telecommunications company.

By any standards it was innovative: it offered remarkable and revolutionary telecom data services, a century before the Internet. But in the century that followed why didn't Western Union become the leader in voice communications, ceding that position to the Bell companies?

Alexader Graham Bell did not invent the voice telephone with the intention of toppling Western Union. In the language of today's technology entrepreneurship he hoped Western Union would be his "exit". He pitched the phone to them as an improvement to their core telegraphy business. Bell offered the patents for the telephone to Western Union for a mere $100,000, roughly $2 million in today's dollars. In other words, today the deal would be: two million dollars to own all voice IP.

Western Union passed

Now why would they do that? Imagine you are an analyst at Western Union advising management on technology acquisition and M&A. What process would you use to evaluate it? First, most obviously, you would try to get feedback from marketing and sales. They should go out and survey their customer base to find out who would be interested and how much they would be willing to pay.

The sales force would naturally seek out their best customers and ask them first. Those customers were the stock ticker users. They were extremely voracious users of the ticker tape system. It should be no surprise why we have "ticker tape parades"--there was enough of the stuff lying around offices in New York to shower down and flood Broadway end-to-end". Ticker tape was a boon to merchants and traders. Chicago would know what's happening in the New York stock exchange instantaneously. Price data was traveling at the speed of light.  The sender and receiver were in lock-step sync. To suggest to these customers a voice product would sound like a step backward. Why would they want to talk to the other party? The data was coming on tape and it was also recorded for posterity. Speaking with the other person meant confusion and delay and a mess of record keeping.

The word from the customers would be a resounding negative.

Second would be the business model.  How would you charge for voice? Data was subscription based, and the fees were huge. Voice would be metered and the price low. If targeting consumers, telegrams were priced by the word, but you couldn't price voice by the word. Marketing would stare back at you as if you had committed a judgement error.

Thirdly you would look at the cost side. Operations would give you more bad news. Voice technology was different enough that it needed new infrastructure. Switchboards, operators, handsets and new power and insulator requirements were all expensive and had no clear return. Voice was a loser all around.

So the decision was perfectly reasonable, smart even.

This is not to suggest that Western Union did not accept any new technology or invest in it. For decades after the phone, Western Union continued to invest in innovation: It introduced an electronic payment system (called money transfer) in 1871. In 1914 it offered the first charge card and in 1923 it introduced teletypewriters. Singing telegrams followed in 1933 and fax service in 1935. Intercity wireless microwave communications were introduced in 1943. Telex came in 1958 and the product of marketing genius, the 'Candygram' in the 1960s. By 1964 its network was all wireless using a transcontinental microwave system. Western Union became the first American telecommunications corporation to maintain its own fleet of geosynchronous communication satellites, starting in 1974. The fleet of satellites, called Westar, carried communications within the Western Union company for telegram and mailgram message data to Western Union bureaus nationwide.

And it wasn't lacking in talent either. Western Union had the best telecom managers in the business and was raking in profits to fund expansion. It attracted the best and brightest engineers for generations. Western Union probably passed on hundreds of other inventions and ideas, and rightly so.

So what happened to WU?

It's still around. However, due to declining profits and mounting debts, Western Union slowly began to divest itself of telecommunications-based assets starting in the early 1980s. Due to deregulation in the US, Western Union began sending money outside the country, re-inventing itself as "The fastest way to send money worldwide". In 1987, an individual investor acquired control of Western Union through an outside of chapter 11 process that was a complex leveraged recapitalization. After several other changes in ownerships and moves through bankruptcy, WU is now a public company with an enterprise value of about $12 billion based entirely on international money transfer business.

3G as a disruptive technology

Today we've come around full circle. Whereas voice was disruptive to a data network provider in the 20th century, data is disruptive to a voice network provider in the 21st. The mis-application of 3G network technology to sustain voice-oriented business models gives one a strong sense of déjà vu.

The telecom world moved swiftly to 1G and 2G cellular voice. Each new generation of mobile telecom was sustaining the core business model of subscriptions feeding centrally managed networks and symbiotic device vendors benefiting from handset and switching system upgrades. Standardization, international expansion, lower price points and network effects made this industry the most ubiquitous technology distributor the world had ever seen.

3G however was different. It was effectively mobile broadband data. It enabled data services to be de-coupled from the network operators. IP protocols did not require the servers to be located on operator premises or even to be managed or monitored by them. Devices could be general purpose computers not specific purpose communicators. Incumbents did not adapt to this well.  They continued to build business plans according to a central-switch blueprint.

So although incumbents did not reject the smartphone as a technology, they tried to make it sustaining when plainly it isn't. They essentially tried to cram it into their business models.

The struggle to keep the service structure of telecom centered around the network operator will continue, but short of pulling the plug on 3G and LTE, there is little that the incumbents can do to stop it.

So what's the lesson?

Some technologies are embraced and some are rejected. You can't fault every rejection and you can't praise every acceptance. The most likely reason for embracing something is that it helps grow your existing business. The most likely reason for rejection is that it may harm your business. Even acceptance sometimes leads to mis-application in sustaining the core rather than planning for its demise.

The decision on how to handle something new and potentially disruptive requires a different sense of what's right and wrong about it.

--

Photo: Former headquarters of WU, located at 60 Hudson, New York.

With 50 million clients installed likely this year, FaceTime is a service which may have potential.  By potential I mean, of course, its potential to make a dent in the universe, the universal goal Jobs has set for Apple so many years ago.

It's entirely possible that FaceTime will be an inconsequential nice-to-have enhancement to the iPhone platform that sustains and improves the iOS (and OSX) franchise. But it's also entirely possible that FaceTime could become a disruptive challenger to the functional voice-oriented telecom network.

What you choose to measure determines what you know

The thing that jumps out at me is the service's under-performance along metrics that the industry defines as critical but over-performance on dimensions that the industry considers non-critical.

It suffers from poor coverage (WiFi only), a limited network of possible users, high bandwidth requirements which are still hard to match and a steep price point for initial entry.

But I would say data-oriented devices were similarly met with poor performance 4 to 5 years ago. No 3G, poor coverage, few services and high prices. Even today there are many telecom skeptics who think smartphones are toys for the well-to-do or IT types and not something for the mass market.  As can be easily seen however, all the shortcomings of the mobile-data-oriented-device (aka smartphone) turned out to get ironed out with time. You just need to plot a historic graph of network performance and device price points to see where they're going to go.

On the other side of the equation FaceTime offers huge performance on the dimension of emotional attachment to a device and to a conversation. The FaceTime experience is qualitatively distinct from either desktop video chat or device-based VoIP. It's intimate, emotional and stirring, (just look at the commercial) not to mention accessible by just about anyone. These new measures of performance cannot yet be quantified with spreadsheets and measuring the effect will not be a priority for anyone in the industry for a long time to come. Therefore, the asymmetry is there implicitly.

So here we have classic Apple: shifting of the basis of competition so subtly that it's almost imperceptible while under way yet glaringly obvious after the fact.

So what does FaceTime have to do with FaceBook?

When FaceBook launched, I don't recall observers predicting that a site allowing college students to keep track of their friends after graduation would grow into the stickiest site by time spent online ever seen. Nor that it would become a site that swallowed all online communications including IM and email. Nor a site that would possibly challenge search consumption. The performance attributes that FaceBook relied on were not in algorithms but in the emotional hooks that relationships have with people. FaceBook was personal, emotive and private, the exact opposite of what the Web offered a decade ago.

So, like the names imply, both these services are about Faces, the first objects we as humans recognized after birth. That's a powerful image to have a handle on.

The pattern that emerges is that Web 2.0, Telecom 2.0 and Personal Computing 2.0 will pivot on the elevation of emotion over function. That may be a banality but empires have been built on less.

Under Kallasvuo Nokia embarked on the most dramatic shift in its business since entering the mobile phone business in the early 90s.

The shift was not into mobile software which began in 2001 under his predecessor. It was not into enterprise solutions which also preceded his tenure. OPK's main contribution was the move into mobile services.

The concept of mobile services may be an unfamiliar one for casual observers because it has not become a visible business for operators and certainly not for handset vendors. It's also a complicated business model that requires some deeper understanding of the way the telecom industry is structured.

What Nokia had in mind was to offer various value-added, billable services to operators which would be enabled by Nokia handsets. The types of services included music subscriptions (Comes With Music), email (several acquisitions), photo sharing, and navigation.

The idea was that since many operators would not be capable of rolling out own brand services and could not do the heavy back-end lifting or the integration with handsets, someone could step in and roll out white labeled solutions world-wide. Third parties would also find it impossible to integrate and would lack the relationships Nokia had with operators world-wide.

For example, Nokia could enable a South American operator to offer email services to all their customers (with or without smartphones) and that service could be offered at a certain incremental price over the basic voice plan. The client implementation could be device independent but Nokia devices would probably work better. This would lead to higher ARPU which could be shared with Nokia.

Anyone can see that this is a complicated business plan and is therefore unlikely to be successful. But what makes it a complete failure is the realization that most buyers will resist the idea of paying for individual services separately. $1/mo for email, $2/mo for music, $3/mo for maps, etc. is repulsive. Users stampeded instead to unlimited data plans and smartphones which offered all these services and hundreds more for free or at prices negotiated with third party providers, rather than the untrusted network operators.

And therein lies the entire cause of Nokia's strategic failure: an operator centric point of view. It led to poisoned devices and irrational business plans.

Which leads to the question in the headline. Is this mistake recognized and is it big enough to cause such a disruptive CEO dismissal?

I argue yes. Strategic errors are forgivable, but the they become a capital offense when they turn into a derailment of the core business. Instead of being enhanced with value-added services, the core business collapsed under the disruptive attack of unlimited data.

But it gets worse. Like the capital offense that Robbie Bach was guilty of at Microsoft, there has to be some direct accountability. To add insult to injury, OPK single-handedly pushed through the biggest and stupidest acquisition in Nokia's history. To support this flawed vision of mobile services OPK bought Navteq for $8.1 billion in October 2007.

Intended as a service that could be rolled out on all phones and monetized through operator billing, Nokia maps is a free service that will never return anything to shareholders.

Missing where the puck was going is one thing but burning precious capital is another. This, in my humble opinion, is why OPK was fired.

Speaking of pucks, here's hoping fresh Canadian eyes will see where it's going.

Verizon, unfortunately, is also what ruins the phone. Or, rather, what it’s forced Samsung to do to the phone, which you could sum up in a word: Bing. Bing is the default—and only—search engine on the Fascinate. A Google Android phone. In the search widget, in the browser, when you press the search button. Bing. No, you can’t change it. There’s no setting for it, and the Google Search widget that you can snag from the Market is blocked (or at least very carefully hidden). Being unwittingly forced into Verizon and Bing’s conjugal relationship is infuriating on its own, but the implementation also feels like the sloppy hack that it is.

via Daring Fireball Linked List: Matt Buchanan on Verizon's Samsung Fascinate Lightning.

John Gruber astutely adds:

Android is “open”, but who it’s open for, primarily, are the carriers. (Somehow I doubt we’ll see any Windows Phone 7 devices where Google is the one and only search option.)

The primary defense of Google's Android strategy is that it's beneficial in driving traffic to Google's services/properties. This is by no means a certainty. To the contrary, it seems likely that the Android experience will be defined by operator back-room deals.

See also: asymco | Android vs. Google Part II

Coupling a lack of control over the platform, the revenue streams, the user experience, the potential banishment of AdMob from iOS and an attack on Google's brand, Android is currently winning the war with Google.

However, my money long term remains with Google. They can and will eventually beat Android. Perhaps with Chrome.

[Footnote: if anyone wonders why Verizon, Google's best friend in mobile, is gutting Android, you need to remember an exclusive five-year deal Microsoft struck with the carrier to provide search and advertising services on the phone. Microsoft was rumored to pay $500 million for the opportunity.]

In 2007, 14% of Britons' online time was spent on IM, according to the UK Online Measurement company - but that has fallen to just 5%, the firm says, basing its findings on the habits of a panel of 40,000 computer users.

The study was released shortly after AOL sold its ICQ instant messaging service $187.5m (£124m) - less than half what the company paid for it in 1998.

And in September 2009, a survey of internet use by the New York-based Online Publishers Association found that the amount of time spent by surfers on traditional communications tools, including IM and e-mail, had declined by 8% since 2003.

via BBC News - Instant messaging: This conversation is terminated.

In other news today Yahoo and Nokia announced a worldwide partnership. Yahoo will provide e-mail and chat services on Nokia phones. The services will be co-branded.

OMG LOL.

So Dell is making Android mobile phones while Nokia is making Windows PCs. A curious confluence of poor management decisions.

Why is all this dilution of core competence happening? Average punditry would suggest it's all about "convergence" and the overlapping of mobile and until-recently-fixed computing.

Er, not quite.

The economics of plastic bits shipped in cardboard boxes has not changed. It's as lousy a business as it has always been, whether the plastic bits are small or large. Some would argue that the gross margins of mobile devices are higher than those of computers, but that's not true on average. The gross margin on average mobile phones is just as low (<10%) as it is on PCs. There are about the same number of competitors and rivalry and global marketplace.

If indeed, margins were the only story, why would Nokia choose to chase the low margin PC business while holding (albeit very tentatively) 50% global share in "smartphones"? Similarly, why would RIMM choose to chase the low end consumer business (link: Blackberry Pearl et. al.) while having the fat margins of thinly penetrated business devices to wallow in? Finally, why would there be hundreds (literally link: www.pdadb.net) Windows Mobile smartphone licensees, the vast majority of which don't manage to sell more than the 50k minimum volume licenses they are obliged to order?

The fact of the matter is that the money being chased by device vendors is not being paid for devices, but for the service plans those devices enable. More precisely, these businesses are concerned with capturing a share of ARPU uplift from mobile data services. As such, they are mostly offering hardware as enticements for the *change* in ARPU that mobile broadband creates. Success depends on the power of that hardware to compel users to switch to higher data service fees.

The first proponent in his "hardware as ARPU booster" was actually Microsoft back in 2003 when they entered the mobile OS market. One of their value propositions was that Windows Mobile would boost data ARPU. The reality was not so compelling however as WinMo turned out to be too awkward to use for most users and the power of the platform lay fallow. In reality, the first successful hardware ARPU booster was RIM. In the middle of this decade, RIM showed that a Blackberry user could be worth between $5000 and $8000 in NPV ARPU to an operator. For that juicy pie, RIM was handsomely rewarded with a small "cut" that boosted their margins to above 50% making them the most profitable device vendor on the planet.

The realization only began to set in about 2007 as every other vendor began applying resources to deliver ARPU boosters. Apple took note early and executed beautifully with the iPhone, mastering the "get paid for ARPU" game, and in the process capturing 50%+ gross margins and (depending on the quarter) up to 30% of all profits in the device industry.

To forecast how this game will be played and how sustainable this boosterism is, we need to segment or categorize the market according to ARPU. The way to visualize this is what I call the "wedding cake" model. Each cake layer represents the population of users in that ARPU segment and the thickness of that layer represents the delta of ARPU to the layer below.

• The base layer is huge (billions of customers) and represents the prepaid voice market. The average "thickness" of this layer is about $10/mo.
• The next layer is smaller and represents the postpaid voice market. The thickness is about $30, representing the "boost" over prepaid.
• The next layer is smaller still and represents the postpaid voice+data market where the data is a-la-carte. This is typical of plans where the buyer pays separately for certain data services (e.g. email, navigation, etc.) and pays on a per-megabyte basis for browsing. The ARPU differential is about $20.
• Finally the top layer is the unlimited data plan packaged with bucket voice and SMS. Here the ARPU differential is another $30.

At its highest, the cake is about $100/mo., though the width of the top layer and hence the population of paying customers at that price point is quite small relative to the other layers.

Now each layer is covered with a "frosting" which is the device hardware revenue. Obviously the frosting gets a bit thicker the higher the layer, but it's a relatively thin layer compared to the "cake" thickness. In the higher layers the trend is in fact to make the frosting as thin as possible (<$99 NPV or <$4/mo. for a 24 mo. contract.) The devices which enable the top layers are clearly more complex and expensive to make, but the end-user pricing is trending down. The trick is therefore to convince the operators to pay a piece of the "cake" to the device vendor in exchange for migrating users to these higher layers.