“Strong opinions weakly held” is a core principle of mine, and while I think I’ve demonstrated strong opinions aplenty on this blog, today it’s time to give credence to the “weakly held” part.
Specifically, I have been wrong about Facebook. I’ve been a bear about their long-term prospects, and now I am a bull.
I stand by all these articles, but for one line from the third bullet point:
Facebook is an app, not a platform.
An app can afford to be prescriptive about the user experience and means of interaction; in fact, the best apps have a point of view on how the user ought to use their service.
Platforms, on the other hand, are just that: a stage for actors (i.e. apps) of the user’s choosing to create a wholly unique experience that is particular for every individual user.
It follows, of course, that no successful platform can be built on advertising. Advertising demands eyeballs; platform success demands the ability to fade into the background as said unique experiences take center stage.
I’m not sure an app is worth $64 billion.
It’s the last sentence that contains the seed of the fundamental error I made in evaluating Facebook. Specifically, I presumed that only platforms were worth billions of dollars.
Facebook, of course, was making the same presumption about needing to be a platform. Much of their efforts in 2007-2009 were focused on building the Facebook platform with the goal of, well, something other than making money, and they were neglecting mobile to do so. When mobile proceeded to become the biggest avenue for growth (didn’t something else launch in 2007?), Facebook seemed to be in trouble. After all, the platform they had spent so much time building wouldn’t translate.
It turns out, though, that the shift to mobile saved Facebook from itself.
First off, mobile apps own the entire (small) screen. You see nothing but the app that you are using at any one particular time.
Secondly, mobile apps are just that: apps, not platforms. There is no need for Facebook to “reserve space” in their mobile apps for partners or other apps.
That’s why my quote above is actually the bull case for Facebook.
Specifically, it’s better for an advertising business to not be a platform. There are certain roles and responsibilities a platform must bear with regards to the user experience, and many of these work against effective advertising. That’s why, for example, you don’t see any advertising in Android, despite the fact it’s built by the top advertising company in the world.
So a Facebook app owns the entire screen, and can use all of that screen for what benefits Facebook, and Facebook alone.
Something like this:
Two of the first three items in my timeline are ads; at one point they take over the entire iPhone 5 screen
You can’t help but see the advertising, which makes it particularly attractive to advertisers. Brand advertising, especially, is all about visuals and video (launching soon!), but no one has been able to make brand advertising work as well on the web as it does on TV or print. There is simply too much to see on the screen at any given time.
This is the exact opposite experience of a mobile app. Brand advertising on Facebook’s app shares the screen with no one. Thanks to the constraints of mobile, Facebook may be cracking the display and brand advertising nut that has frustrated online advertisers for years.
Facebook made $655.6m from mobile ads between April and June – more than 41% of the $1.6bn in overall advertising revenue, compared with 30% for the same period last year. The number of mobile users expanded 51% to 819 million.
“I’d say that Facebook is in a unique position with the ad market right now in that it is very far ahead of any other major ad platform in mobile display advertising,” said Clark Fredricksen of research firm eMarketer.
The key, of course, is the iron grip Facebook has on its userbase. You know how readers of this blog feel about Twitter? That’s how normal people feel about Facebook. It’s the first thing they check in the morning, on the bus, while getting coffee, and on the can. It is the most indispensable tech product in most people’s lives, and every time one of those billion people use the mobile app, they see an advertisement that completely owns their device’s screen, if only for a moment.
The connection to TV and print are clear: great content draws eyeballs, and the advertising is placed where you can’t miss it. Combine this advertising opportunity with Facebook’s decent signaling, and you have the makings of a single-product company on its way to owning advertising on today’s emerging platform.
That’s the story of Google and the web in 2005, and it’s the story of Facebook and mobile in 2013. And none of us – and I think that includes Mark Zuckerberg1 – saw it coming.
I’m actually very impressed with Zuckerberg and think he is a very good executive; however, Facebook Home is all the evidence I need on this point ↩
The surest route to befuddlement in the tech industry is comparing a vertical player, like Apple, with a horizontal one, like Google.
Vertical players typically monetize through hardware, only serve a subset of users, and any services they provide are exclusive to their devices. Horizontal players, on the other hand, monetize through subscriptions or ads, and seek to serve all users across all devices.1
In Services, Not Devices, I argued that Microsoft’s focus on vertical devices was distracting them from the horizontal services world they are much more suited for. Or, as Narain Jashanmal put it:
In other words: Microsoft, stop trying to compete with Apple & Google, focus on competing just with Google: http://t.co/nJQr6suHY2
Google certainly isn’t confused; they are 100% a horizontal player,2 and fully absorbing the implications of that fact gives necessary color to yesterday’s announcements.
If you recall back to Google I/O, we talked about the fact that we are living in a pivotal moment in computing. Users are increasingly adopting computing devices. Smartphones have exploded in growth. Tablets are following the same trend. The combination of smartphones, tablets, laptops, televisions, it’s a multi-screen world for our users. And increasingly, there’s this whole new category of emerging devices on the horizon. It’s a very very exciting time.
We at Google are incredibly excited by this trend and are embracing it. Our goal is to deliver an experience that is seamless, consistent, and beautiful across all these screens. And the way we do that is by investing in two platforms. Android, and Chrome.
For a horizontal player, the focus in a “multi-screen world” needs to be on all verticals, not just one.
Chrome is the focus at Google; Android is an afterthought
For Google, it is Chrome that fits this focus on a multi-screen world. Chrome shouldn’t be thought of as a web browser; rather, it’s an optimized bi-directional delivery vehicle: the best experience with Google services for users, and maximum user data for Google. And it runs everywhere.3 This is why Google has been investing millions of dollars in building the Chrome brand for some time now.
Android, on the other hand, enables several of those verticals, and keeps Apple honest in phones especially; however, by virtue of the hardware world it lives in, it’s not the best vehicle for reaching all users, and Google is fine with that.
Google isn’t that interested in phones anymore
The only notable thing about Android phones at yesterday’s announcement was their absence.
I’m not surprised. As I wrote after Google I/O, which didn’t focus on phones either, Android was defensive:
Google acquired Android in 2005 as a defense against Windows Mobile dominating smartphones just as Windows dominated PCs. When the iPhone arrived in 2007, Google quickly pivoted Android to defuse the new threat. And they were hugely successful…
For Google, Android was a detour from their focus on owning and dominating web services; it ensured that those services would be freely accessible in this new world of computing, including on the iPhones and iPads that were used liberally in nearly every keynote demo. And, now that Android is successful, Google is back to focusing on [services].
One could make the argument that Google can no longer control Android. I would contend they don’t even want to. In fact, that was the point. No one company will ever control mobile (or a great many other things that will run some variant of Android), but all mobile devices will access the web.4
So, now that Android is good enough on phones, there simply isn’t any point in investing in it as heavily as before.
“Named” Android releases over time
Google is worried about the iPad dominating tablets
The Android team is, however, still investing in tablets. The first half of yesterday’s event focused on the Nexus 7, the launch vehicle for Android 4.3 (still Jelly Bean). A few asides on the Nexus 7:
When horizontal companies build hardware, they almost always sell it at cost. After all, the whole point is usually to enable as many people as possible to access their services
I’m actually not certain the Nexus 7 means there will be a retina iPad mini. The new Nexus 7 costs $229; presuming it is sold at cost, that equates to $327 with a 30% margin – i.e., almost the exact cost of an iPad Mini. However, Apple has hinted that the iPad mini may have lower margins than 30% (meaning its cost-of-goods is higher than the Nexus 7), and a retina mini would almost certainly have higher quality components, particularly when it comes to the screen. $329 will be a tough price to hit this year (note too, that Google had to raise the price of the new Nexus 7 by $30 as it is)5
Google is much more successful, relatively speaking, selling Nexus tablets than they are Nexus phones; they sell 10% of all Android tablets, but a tiny fraction of Android smartphones. I don’t think this is a surprise; the go-to-market in smartphones is much more complex due to the involvement of carriers. Google isn’t great at dealing with such complexities, and given the fact they’re a horizontal player, aren’t that motivated to get better. Tablet sales, however, are much more transactional.
That final point, however, works against Android in tablets. Device makers own the customer relationship and value proposition in tablets to a much greater extent than they do in smartphones, which magnifies Apple’s lead in these areas. Google certainly would love to see the tablet market look more like the smartphone one, but they’re just not that comparable.
Chromecast is an obvious product
Everything I’ve described makes the choices behind Chromecast clear:
Google believes that the future is multi-screen, and the most prominent screen in most consumers’ lives is the television
As a horizontal company, Google wants to be on every screen, and their vehicle to accomplish that across verticals, both from a technical and brand perspective, is Chrome
Chromecast works on all devices – including iOS – not just Android
Chromecast is priced as low as it can be
Chromecast has both a relatively easy go-to-market as a standalone device as well as allure for television makers
It helps that Chromecast looks to be very well-executed – everything Google TV wasn’t, in every way.
As I wrote in The Google We Always Wanted, Google is laser-focused right now. They have a clear identity as a horizontal services company, and from that perspective, just about everything they are doing makes sense.
The economics behind vertical and horizontal business models are interesting; I plan to explore them in a future post ↩
Well, except for that bizarre Motorola thing. My best guess is that the decision was ultimately a combination of misguided kingdom-building by Andy Rubin, patent panic, and delusions of Steve-Jobs-esque grandeur from Larry Page. Give them credit for treating it as a sunk cost (and by sunk cost, I don’t mean abandoned; rather, they are treating Motorola pretty rationally, not trying to justify $12 billion) ↩
Many of those devices, especially in China, have non-Google services attached; since I believe Android was defensive, I also believe that Google is OK with that ↩
To be clear, the decision has already been made; any pundit that suggests this year’s iPad mini will be in any way influenced by the Nexus 7 is an idiot ↩
In recent months, Apple has asked for prototype smartphone screens larger than 4 inches and has also asked for screen designs for a new tablet device measuring slightly less than 13 inches diagonally, they said. The current iPhone 5 has a four-inch screen, while the iPad has a 9.7-inch screen. The iPad Mini, a stripped-down version of its tablet computer, has a 7.9-inch screen.
I can’t wait to buy one.
The idea of thin and fat clients has been around for years.
Thin clients rely on a central computer for processing; fat clients do most of the processing themselves
A thin client (sometimes also called a lean or slim client) is a computer or a computer program which depends heavily on some other computer (its server) to fulfill its computational roles. This is different from the traditional fat client, which is a computer designed to take on these roles by itself…
Thin clients occur as components of a broader computer infrastructure, where many clients share their computations with the same server. As such, thin client infrastructures can be viewed as providing some computing service via several user interfaces. This is desirable in contexts where individual fat clients have much more functionality or power than the infrastructure requires.
Thin-client computing is also a way of easily maintaining computational services at a reduced total cost of ownership.
While thin client architectures have gained some traction in certain line-of-business applications, the idea has never really taken hold broadly, and certainly not in the consumer market.
The reality of computing over the last 25 years has been a constant improvement in performance coupled with a constant decrease in price. Moreover, for the first two decades of the PC, there was a noticeable benefit for most users in that increased performance. The result was that the reduced performance of thin clients wasn’t justified by a minimal price advantage that was constantly disappearing.1 And so, for the vast majority of uses and the vast majority of customers, fat clients ruled the day.
Or rather fat client, in the singular. For 25 years users typically relied on one general purpose fat client PC that was capable of accomplishing a wide range of tasks. The beneficiaries, of course, were Microsoft, who owned the dominant fat client operating system, and Intel, which provided the ever-improving performance.
That advantage is famously slipping away – or rather, falling off a cliff – in the face of tablets, (literally) thin devices with a new input method – touch – and new operating systems designed for different use cases and constraints.
The result is not a displacement of PCs, but rather an augmentation. The vast majority of people2 who have tablets also have a PC, albeit one that they likely purchased a while ago. Performance finally ended up being good enough, and money was better spent on all the new uses unlocked by tablets.
The result has been a split in people’s computing – and tablets have taken the majority.
Most people now use two computing devices, and tie them together with the cloud
Still, the experiences aren’t entirely disparate; they’re still tied together by the cloud. Perhaps it’s as basic as email being in sync across devices (phones too, of course), or perhaps the user is more sophisticated and uses something like Dropbox. Regardless, their computing experience is not completed by just one device, and not everything that matters happens locally.
In a way, it’s a thin architecture.
To be clear, the classic definition of a thin architecture dictates that processing happens on the server, and that’s mostly not the case today. Tablets are very capable machines with full app libraries and modern browsers that run client-side code with aplomb. But that’s actually kind of the point: maybe the reason thin architectures never took off was that we had the definition – and associated advantages – all wrong.
While a classic thin architecture moves processing to the server, enabling cheaper clients, those clients still have a GUI, mouse and keyboard. In other words, the experience is largely the same as a fat client, minus the superior performance and responsiveness. Tablets, however, are orthogonal to PCs; they are inferior in some ways (performance, text entry), but superior in others (size, battery life, touch). They have a reason-to-own other than price.
Thinking about capabilities beyond processing casts Microsoft’s Windows 8 troubles in stark relief. Windows 8, with it’s mixture of touch and WIMP-interface is the ultimate fat client. But by combining so many capabilities, it necessarily compromises them as well.
Today’s thin clients, on the hand, specialize. A pure tablet is superior for touch-based applications; a pure PC is superior for keyboard-and-mouse ones. An e-ink reader is superior for reading, and a 13-inch iPad would be superior for (in my case) drawing and making music. And while many people now use two devices, I think that’s only the beginning (I’m personally at four and the 13″ iPad would be number five).
Two devices is only the start. Many people, including me, already use multiple. Also, this was very hard to draw on an iPad mini!
PCs are like Humpty Dumpty; they are being broken into pieces and won’t be put back together again.
Incidentally, this squeeze is the story of feature phones in the face of the Android onslaught ↩
At least in the developed world; the developing world is another very interesting story entirely ↩
The larger question remains, though: why did Microsoft reorganize, and what should they have done instead?
“Nothing is more important at Microsoft than Windows.”
— Steve Ballmer, CES 2012 Keynote
While last week I praised Microsoft’s diversification, from a profit-standpoint Microsoft is a three-product company: Windows, Office, and Server, and the latter two are direct beneficiaries of Windows dominance. More Windows machines mean more opportunities to sell Office. More Windows machines with Office mean more opportunities to sell Windows Server and its associated products like Exchange and SQL Server. Steve Ballmer is no dummy: he built Microsoft’s sales side, and knows exactly how critical a solid Windows base is to every other profit-generating product Microsoft sells.
Unfortunately, Windows is foundering. Division profits fell by 54% last quarter (8% if you exclude the $900 million Surface charge), and PC shipments are falling off a cliff. If you believe, as Ballmer likely does, that this imperils all of Microsoft, then there is a certain logic to putting all of the company’s resources to work to prop up Windows. From the One Microsoft memo:
We are rallying behind a single strategy as one company — not a collection of divisional strategies. Although we will deliver multiple devices and services to execute and monetize the strategy, the single core strategy will drive us to set shared goals for everything we do. We will see our product line holistically, not as a set of islands. We will allocate resources and build devices and services that provide compelling, integrated experiences across the many screens in our lives, with maximum return to shareholders…
We will reshape how we interact with our customers, developers and key innovation partners, delivering a more coherent message and family of product offerings. The evangelism and business development team will drive partners across our integrated strategy and its execution. Our marketing, advertising and all our customer interaction will be designed to reflect one company with integrated approaches to our consumer and business marketplaces.
The crux of the problem is in that second paragraph: no one is asking Microsoft to design its “customer interaction” to “reflect one company.” Customers are asking Microsoft to help them solve their problems and get their jobs done, not to make them Microsoft-only customers.
The solipsism is remarkable.
The truth is that Microsoft is wrapping itself around an axle of its own creation. The solution to the secular collapse of the PC market is not to seek to prop up Windows and force an integrated solution that no one is asking for; rather, the goal should be the exact opposite. Maximum effort should be focused on making Office, Server, and all the other products less subservient to Windows and more in line with consumer needs and the reality of computing in 2013.
Devices are vertical, services are horizontal
The trouble for Microsoft in the devices layer is that they only know horizontal domination. When there was nothing but PC’s, the insistence on one experience no matter the hardware worked perfectly. However, a Dell and an HP are much more similar than a tablet and a web page, for example, each of which has its own input method, user expectations, and constraints. A multi-device world demands bespoke experiences, not one size fits all. Microsoft simply doesn’t seem to understand that, and the longer they seek to “horizontalize” devices the greater the write-offs will become.
However, look again at that picture: there remains a horizontal layer – services – and it’s there that Microsoft should focus its energy. For Office and Server specifically:
Documents remain essential and ubiquitous to all of the world outside of Silicon Valley; an independent Office division should be delivering bespoke experiences on every meaningful platform. Office 365 is a great start that would be even better with a version for iPad
A great many apps are simply front-ends for web-based services; an independent Server division should be delivering best-in-class interfaces and tools for app developers on every meaningful platform
As for Windows, let it focus on solidifying Microsoft’s hold on the enterprise (it’s here the need to fight the iPad is most acute), with a nice spillover into Home PCs and gaming, and accept the fact Windows was only ever relevant in the consumer market because nobody got fired for buying IBM.
“Devices and services” is only half right; unfortunately Ballmer picked the wrong half.
Consider this Part 2 in an accidental series on Microsoft’s recent reorganization, and functional and divisional organizations. Part 1 focused on a divisional organization, while today’s Part 2 focuses on functional ones.
The “Uncanny Valley” is most typically associated with animated films (although it was originally about robots). From Wikipedia:
The uncanny valley is a hypothesis in the field of human aesthetics which holds that when human features look and move almost, but not perfectly, like natural human beings, it causes a response of revulsion among human observers. Examples can be found in the fields of robotics, 3D computer animation, and in medical fields such as burn reconstruction, infectious diseases, neurological conditions, and plastic surgery. The “valley” refers to the dip in a graph of the comfort level of humans as subjects move toward a healthy, natural human likeness described in a function of a subject’s aesthetic acceptability.
The graph looks like this:
Hypothesized emotional response of human subjects is plotted against anthropomorphism of a robot, following Mori’s statements. The uncanny valley is the region of negative emotional response towards robots that seem “almost human”. Movement amplifies the emotional response — Wikipedia
I actually think the concept can be applied to many more things than just human aesthetics: anything that is almost perfect, but not quite, is actually worse than something that is good enough:
Consider organizational design. While I am very critical of Microsoft’s decision to reorganize functionally, that does not mean I think that functional organizations are bad. Quite the contrary, in fact. I (and Apple) believe that functional organizations are an essential ingredient to producing truly great products. They are the “best.”
But the reasons functional organizations are theoretically “best” are the exact same reasons why it is so difficult to manage a functional organization. Get any of the following wrong and your organization – e.g. Microsoft – is in a far worse place than if you had settled for a “good enough” divisional structure.
Functional organizations rely on true collaboration
Collaboration is one of those buzzwords that will get you wasted in a drinking game; it’s on the lips of every innovation consultant, thought leader, and CEO. The concept – working together to make something great – seems straightforward, but the fact it’s always the proposed cure is evidence enough it’s hard to achieve.
The reason why is easier to understand when you understand what goes into true collaboration: mutual trust and respect, and the willingness and freedom to disagree.
Both ingredients are really hard for humans in groups. Start with mutual trust and respect: we’re all naturally suspicious of those we aren’t intimate with, and that’s compounded by the anxiety of status and salary inherent in a workplace. It’s even worse at top companies, where nearly every employee grew up the smartest kid in their class and can’t help but try to prove they’re still the best.
As for disagreement, it’s actually the willingness that is more challenging than the freedom. It turns out we humans usually want other people to like us. In fact, we usually want other people to like us more than we want to make something great, so we hesitate to criticize and point out flaws.
The upside should be obvious: it really is possible for the whole to be greater than the sum of the parts. As much as we glorify the solitary coder or designer, the truth is that, especially in technology, breakthroughs are collaborative and iterative, borne of thousands of disagreements. It’s one reason Apple is absolutely committed to a functional organization; there, design, software, hardware, operations, and product marketing are all separated; every decision is necessarily made collaboratively, and (theoretically) results in a better outcome.
However, a company organized functionally that does not have the necessary cultural ingredients for true collaboration is actually in worse shape for having tried; the disagreements will turn into conflict, and the agreements made through groupthink will arguably be worse.
As I noted last week, saying that Microsoft lacks “the necessary cultural ingredients for true collaboration” is an understatement.
Functional organizations rely on effective, visionary leaders
Functional organizations are unusually reliant on a visionary leader. Look again at the graphic of a functional organization:
A functional organization is organized by function. Products are matrixed.
Everything is tied together at the CEO-level, which means communication channels must be excellent, both from the bottom up and also across functions. Moreover, the vast majority of employees only ever see a piece of the product; it’s up to the CEO to set the broad vision and ensure that it is being followed.
The upside is a holistic and integrated quality to the resultant products that is truly extraordinary. People talk about how the iPhone integrated hardware and software, but that is just a crude approximation for what is effectively a single vision of what a phone should be, down to the tiniest detail.
It’s on this point, actually, that concern and criticism of Tim Cook is most warranted. I’ve already made the case as to why he is a great CEO for Apple, but his greatest weakness – product vision – is hugely magnified by the fact that Apple is organized functionally, and thus especially reliant on a visionary leader (this should, of course, make it obvious why Steve Jobs was so insistent on a functional organization). Clearly Jony Ive has been nominated to fill the role of visionary; it is likely that at least one of Ive’s once-peers was not happy about that elevation.
As for Microsoft having a visionary leader, well…
A divisional org, on the other hand, pushes communication and decision-making down into the hierarchy; this makes broad, company-wide endeavors less likely – both breakthroughs, and disasters.
Functional organizations are focused
One of the more interesting elements of the Microsoft reorganization is that it in some ways seems inspired by former Windows President Steven Sinofsky and his thoughts on One Strategy. Sinofsky was a major proponent of functional organizations, and in fact transformed the Windows division from a divisional structure to a functional one, at least on the engineering side.
Both Windows 7 and 8 were the products of that organizational change, and bear out the idea that focus and prioritization make all the difference in the world (leaving aside the merits of the Windows 8 design, it is an engineering marvel, especially considering the timeframe within which it was built).
Still, the fact that Windows is such a massive product meant that Sinofsky himself estimated he only saw a fraction of what mattered:
I have been asked by many what it is like to be a “business leader” and I remind people that even at my job I have no say in tons of things that matter a lot to whether our work is successful (when thought of from the 4 P’s of marketing: product, price, promotion, placement I only have say in 25% of the business, at best). But we have built our team, both in terms of feature teams and in terms of engineering structure so that we can have a balance if we all operate from a shared plan.
If Sinofsky, one of the most iron-fisted executives you will ever see, only saw 25% of what mattered despite sitting at the only nexus for information flow (as noted above), what chance does Ballmer have sitting at the intersection of Windows, Office, Service, Xbox, Bing, etc.?
This is something for Apple to be concerned about, as well. Things like TVs and watches are fun to talk about, but they exact a real cost on the leadership team’s ability to manage the organization.
Functional organizations rely on experts
This is the necessary bookend to collaboration, one of the biggest reasons that functional organizations produce better products. If you are creating something new, would you rather its creators be the best in the world at what they do, or that they be someone trying their hand at a new role with an eye towards career advancement?
I touched on this briefly last week, but it’s worth a closer look:
The ideal employee in a functional organization is an expert who truly loves the type of work they do. Their reward is to do more of what they do, or perhaps lead others who do the same type of work.
The ideal employee in a divisional organization is a generalist. They may start out doing a particularly type of work as an individual contributor, but their next job is likely to be about broadening, not deepening, their skills
There are thousands of employees at Microsoft building nice careers as well-rounded managers; they are all now (theoretically) superfluous, as broad management is only practiced at the CEO level in a functional organization. All these middle managers are likely to either sabotage the reorganization by clinging to their former authority, or simply leave the company. While the latter may be a theoretically good outcome, that is only the long run, in which time we are all dead. In the short run it could be a serious talent and experience drain.
The challenge for a functional organization is more subtle. The best sort of employee finds fulfillment in nothing more than the work they do; the more typical one eventually seeks greener pastures and/or burns out.
As I noted last week, all companies start out functionally organized. After all, they only have one product.
So why do they change, especially if functional companies make better products?
Let’s follow the typical path: Company A makes an amazing product, finds a great market fit, and starts to make a lot of money. They IPO. They continue to grow, and the stock goes up. And then the stock stops going up, because it’s not clear how they will continue to grow. A stock’s worth, after all, is simply the discounted sum of future earnings.
And so the company looks for another avenue of growth. They diversify, maybe successfully, but now they have two products. And soon, like DuPont, they see the wisdom in having two divisions.
Of course, those divisions are certainly related in some way, and it’s inevitable that considerations are given – or dictated, from the CEO – that decisions in one divisions favor the other division whenever possible. This consideration is called a strategy tax, and it’s a hindrance to product quality. So is the inevitable competition for resources, and the increasingly divided attention of the CEO.
It’s all interconnected. There is a tension between product quality, growth, and the stock market. The problem for Microsoft is they have already made their choice: they are a diversified company with a clear focus on the top and bottom lines. It’s not that a functional organization is a bad thing – it’s a great thing – but Microsoft is simply no longer capable of climbing that hill.
DuPont, the famous chemical company, was actually built on gunpowder.1 Founded in the early 1800s, DuPont was a small family concern until the early 1900s, when Pierre DuPont modernized and organized the company around functions: primarily sales and manufacturing. The structure served DuPont well, particularly in World War I, when in response to overwhelming demand DuPont vertically integrated its supply chain, and grew to become one of the largest companies in the world.
After the war, DuPont needed to diversify, and paint, which involved a similar compound to gunpowder, was the area they chose to focus on. Yet, despite the fact DuPont was perhaps the most professionally run corporation in America, losses soared. Eventually, a disconnect between sales and manufacturing was identified as the root cause, and the cure was a new organization around two separate gunpowder and paint divisions.
And thus, the divisional structure was borne. It returned DuPont to profitability, and remains the model for nearly every corporation of significant size, except, notably, for Apple.
And now, Microsoft.
Steve Ballmer restructured Microsoft yesterday as a functional organization. The immensity of this change can not be understated, nor can the risks. Ultimately, I believe the reorganization will paralyze the company and hasten its decline.
The differences between a functional and divisional organization are simple, but profound.
A divisional organization is organized by product.
In a divisional organization, different products are companies unto themselves. They have their own marketing, their own engineering,2 and their own finance. There may be some centralized functions, such as legal and HR, but everything that makes money for a product lives within that product’s organization.
Most crucially, each product has its own profit-and-loss statement (P&L). The performance of each division is thus clear to everyone from the CEO to the division presidents to Wall Street, and accountability is usually directly tied to the P&L. So is motivation: senior division leaders have much of their income tied to the performance of their division.
This model scales to a very high number of products. Consider General Electric, the classic example of a divisional company. It has twenty-five different businesses, ranging from finance to jet turbines. The competitive advantage of such companies is usually in their management acumen and capital reserves, and the preferred employee is a generalist, able to quickly master any job with a refined set of skills.
A functional organization is organized by function. Products are matrixed.
Functional organizations are the exact opposite: each function is siloed, and products cut across functions in a matrix-like fashion. This significantly expands the role of the CEO and his leadership team in all products, as that is the main point of coordination.
In this model, there is usually no ownership for a product’s P&L. Instead, everything accrues to the company’s all-up P&L, including compensation. In addition, the ideal employee is a specialist, not a generalist.
Most functional organizations are single-product companies; to put it another way, almost all companies start as functional organizations. In fact, the moment a company switches to divisions is as good a marking point as any for the end of startup-hood.
Apple is the exception that proves the rule; they are functionally organized, but that absolutely does not mean a functional organization is best. Understanding why almost all corporations are organized by divisions (and how Apple manages a functional organization) illustrates why I think Ballmer is making a grave mistake.
Divisional organizations are better suited to multi-product companies
Microsoft has 13 distinct products and services with a run-rate of more than one billion dollars a year: Windows, Office, Xbox, SQL Server, System Center, Unified Communications, SharePoint, Developer Tools, Dynamics (ERP & CRM), Online display and search advertising, and Azure.
First off, that’s incredibly impressive and is one of Ballmer’s greatest accomplishments.
Second, look again at that list. There are certainly commonalities; for example, all of the products are built on software, just like DuPont, where gun powder and paint were both based on nitrocellulose. But the markets for Xbox and SharePoint, say, are totally different, again, just like they were for gunpowder and paint.
Multiple products in a functional organization are a recipe for a complete disconnect between functions, and Ballmer is putting at significant risk his greatest legacy: Microsoft’s diversification. Smaller, yet growing divisions will be hurt by this change.
Apple is the counterexample here; a central tenet of management’s thinking is that Apple only ever have a small number of products. This isn’t just about design; Tim Cook is absolutely cognizant that a functional organization cannot effectively manage too many products.
Divisional organizations have built-in performance incentives
It’s easy to ignore talk about P&L’s and performance incentives, but the truth is few things are more critical to how organizations are run. People do what they are incentivized to do; it follows that employees who are compensated based on the performance of their division are particularly motivated to ensure the success of the product they work on.
Of course, compensation-based motivation is far inferior to mission-based motivation. For Microsoft, that mission used to be clear: “A computer on every desk and in every home, running Microsoft software.” Today, not so much. From Ballmer’s strategy memo:
Going forward, our strategy will focus on creating a family of devices and services for individuals and businesses that empower people around the globe at home, at work and on the go, for the activities they value most.
That’s so bland as to be meaningless, which means the impact on Microsoft’s shift to a functional organization are profound: what is going to motivate employees working on a huge array of products whose contributions are so far removed from the all-up P&L?
Again, the Apple counterexample: I wrote a few weeks ago that the latest Apple commercial – Designed in California – was written for Apple’s employees. This is why. Apple is uniquely reliant on non-monetary motivation; their employees genuinely believe they are making products that impact people’s lives. I doubt the same could be said for a random marketer on SQL Server.
Divisional organizations have built-in accountability
The flipside of motivation is accountability: the fact each product has its own P&L makes it very clear which division leaders are succeeding or failing.
Things are much less clear in a functional organization. To go back to the DuPont example, when the business was failing fingers were originally pointed at Sales. It was only when the vice-president of sales successfully demonstrated that they were doing nothing wrong that the company searched for another root cause.
Microsoft will face the exact same problem. Engineering will blame marketing who will blame finance who will blame engineering. This will absolutely be exacerbated by Microsoft’s internal culture, which steadfastly glosses over bad news and quietly ships out underperformers while throwing parties and praising their contributions.
This is something Apple is very clear about, stressing the idea of accountability without responsibility – you may be dependent on others, but if you fail, it’s your fault. Thus Rob Schoeben was fired for MobileMe, Mark Papermaster was fired for the iPhone 4 antenna, and Richard Williamson was fired for Maps. I have not seen any evidence of the same sort of ruthlessness at Microsoft recently, even with the clarity of divisional P&Ls.
Divisional organizations have clear career advancement opportunities for generalists
I’ve already noted that divisional organizations favor generalists; generalists also tend to favor divisional organizations as there is a much clearer career path. You are able to move up not only within a specific function, but you can also move laterally amongst divisions, usually with a nice jump in responsibility.
Microsoft has been hiring this type of person for years now: career-focused, monetarily-motivated generalists. In other words, the exact sort of person who does not do well in a functional organization that values expertise and experience.
I interned at Apple as an MBA student, and while I had a fantastic experience, it was well-known that MBA’s typically did not last long at Apple due to the limited options for career advancement and the fact you were only ever hired for what you did previously, not for the broader expansion of skills you supposedly picked up in school.
If Microsoft is truly committed to a functional organization they are likely going to face employee backlash, particularly on the business side.
Divisional organizations have competitive cultures; functional organizations need collaborative ones
Here I think we’ve finally arrived at the motivation for the switch.3 Cross-group collaboration at Microsoft is famously difficult:
Comic from Bonkers World
However, if this is indeed the goal, reorganizing the company to improve collaboration is like burning down your house to kill termites. It ignores the root cause and a much more effective solution.
Last fall Kurt Eichenwald wrote a must-read piece about Microsoft’s lost decade. Every part of it was spot-on, particularly the section on employee reviews and stack ranking:
At the center of the cultural problems was a management system called “stack ranking.” Every current and former Microsoft employee I interviewed—every one—cited stack ranking as the most destructive process inside of Microsoft, something that drove out untold numbers of employees. The system—also referred to as “the performance model,” “the bell curve,” or just “the employee review”—has, with certain variations over the years, worked like this: every unit was forced to declare a certain percentage of employees as top performers, then good performers, then average, then below average, then poor.
“If you were on a team of 10 people, you walked in the first day knowing that, no matter how good everyone was, two people were going to get a great review, seven were going to get mediocre reviews, and one was going to get a terrible review,” said a former software developer. “It leads to employees focusing on competing with each other rather than competing with other companies.”
In my (very-biased) opinion, I believe collaboration is fundamentally broken at Microsoft. It is all about politics, not great outcomes, and that is absolute death in a functional organization, which has nothing but collaboration to hold together cross-functional product teams. At least in a divisional model all of the relevant team members have a common product and a common boss, meaning everyone has no choice but to work together. Unless the employee review and compensation model is significantly changed, this, along with the lack of mission and clear accountability, will grind progress to a halt.
I harbor no ill will towards Microsoft. They were a great employer and I loved my team, annual reviews notwithstanding. However, the impetus behind this move seems sadly characteristic of many of Microsoft’s initiatives – great on the surface, not so great in the details.
In this case, on the surface, functional organization are collaborative ones, and Microsoft needs to collaborate better. However, a closer examination as to why almost every organization of any significant size is divisionally organized suggests this may be Ballmer’s final folly.
Most of the DuPont facts are drawn from Richard Tedlow’s excellent book, “Denial: Why Business Leaders Fail to Look Facts in the Face – And What to Do About It.” ↩
Or manufacturing, or whatever the primary function of the company is ↩
Although I’m not ready to dismiss pure Apple envy ↩
Benedict Evans, in Glass, Home and solipsism, one of the most insightful posts I’ve read in some time:
Your customers’ relationships with you are the only relationships you have as a business and you think a lot about them. But you’re one of a thousand things your customer thinks about in a week, and one of dozens of businesses. And they probably have their own ideas about how they want to engage with you (though they wouldn’t put it in those words) – assuming they think about you at all.
This applies even to Google or Facebook (which brings me to the title of this post). There’s lots of data showing the high proportion of online time that people spend using Facebook, and the high volume of web searches that they do using Google. Facebook and Google are important. But that doesn’t mean they’re everything.
It applies to Apple too, and there’s no better example than iCloud versus Dropbox.
Steve Jobs famously called Dropbox “a feature, not a product.” And, if you look at the Mobile Hierarchy of Needs, he was right; Dropbox sits firmly in “Services,” one piece of the overall device value stack.
The Mobile Hierarchy of Needs – this view supposes one device
But devices – even Apple ones – are ultimately a means to an end. For my mother, her iPad lets her see pictures of the grandchildren; for my wife, her PC was a lifeline to Taiwan for four years; for me, my phone keeps me connected and engaged in technology, regardless of time or place.
It’s the ends that are meaningful and infinitely valuable, and Dropbox’s approach to my most important data1 is much more in line with the value I ascribe to that data: it’s available everywhere.
Not so for iCloud: data is available only on Apple devices, and it’s not exactly clear how to get it out. iCloud is beneficial to Apple’s business model – more devices – yet actively hostile to my needs, in much the same way Facebook Home ignored how people actually use smartphones.
The problem for iCloud, and by extension Apple, is that services – driven by the need for data to be everywhere – are a horizontal layer, orthogonal to the the vertical nature of devices.
Dropbox and other services are a threat to Apple in that they ultimately devalue hardware and minimize switching costs; this is exactly why Apple won’t buy Dropbox (except to kill it), and will never build a fully comparable product.
The only coherent strategy for Apple is a walled-garden of sorts that protects their vertical business model. A services-centric company like Dropbox, on the other hand, ought to pursue a horizontal strategy predicated on maximizing the number of interconnects with the layers above and below; clients on every platform address the latter, and yesterday’s announcement of the Dropbox Platform are focused on the former.
Today, at Dropbox’s first-ever developers conference, the company is officially launching a new set of coding tools designed to push Dropbox into every corner of your digital life. Not content to stay sequestered inside the box, the company’s co-founders are unveiling ways for developers to meld their service with every app on every device you own…
Dropbox may be in a unique position to do just that. For now, the company has no overarching obligations to Apple, Google, Facebook, or Microsoft. While pundits have recommended that some big tech company ought to snap up Dropbox while it’s still relatively affordable, Dropbox is in a position where it can layer itself over competing operating systems while being beholden to none. Says Houston: “No engineer in Cupertino is thinking: ‘How do I make this work with Android?’” Leave that thinking to Dropbox…
Taken together, she says Datastores and Drop-ins transform Dropbox into a platform that enables a “pervasive data layer” — a way for all your digital stuff to follow you everywhere, regardless of device, operating system, or app.
The strategy is spot-on; while iOS and Android (and mostly Samsung) have effectively won the battle for devices, the winner in horizontal services will certainly not be Apple.
Google has pole position. Android was a detour to ensure the services layer was competitive, and now Google is fully focused on winning. However, Google is not as developer friendly, and the way they view data as a commodity can result in user-hostile choices. Moreover, there is always the potential of Google favoring Android, although I think the removal of Andy Rubin was about making sure this doesn’t happen
Microsoft has more strengths here than most people realize. Skydrive is one of the best products they make, and Azure cloud services already provide most of what Dropbox is rolling out. The challenge for Microsoft is perception and mindshare amongst developers especially. Moreover, Microsoft’s strategy of devices and services is about competing both vertically (devices) and hoizontally (services). This is a great recipe for strategy taxes
Amazon has pieces, particularly cloud storage, and they are clearly willing to compete on price. However, to say that Amazon struggles with great user interfaces is putting it kindly, and I don’t see any real evidence they are interested in anything they can’t sell
Yahoo! has massive reach, and some intriguing pieces, including Mail, Flickr, and now Tumblr. However, they are pursuing the Google signal-to-ads monetization model, and haven’t been relevant to developers in…ever? Yahoo should have bought Dropbox
Dropbox has a great base product, a monetization model that is aligned with the value customers place on their data, and excellent mindshare amongst app developers especially. However, Dropbox has a much more limited reach, and I’ve seen precious little evidence they have the marketing chops to overcome that
There likely won’t be one clear-cut winner, as the individualized nature of data works against the network effects that would lead to winner-take-all. That said, Dropbox’s developer mindshare is a massive asset in this regard, especially if Dropbox is able to marshall developers as Dropbox advocates.
Regardless, that set of contenders is good company and gives a hint as to why Dropbox’s valuation is so high.
I’ve been a paying customer for going on six years ↩
Box, for example, makes noises about consumers but is an enterprise company; Dropbox is the opposite ↩
Ahead of the App Store’s fifth anniversary on Wednesday, July 10, Apple has launched a new promotion that includes five “groundbreaking” iOS apps and five “landmark” iOS games; these apps will be available for free for a limited time to celebrate the first five years of App Store. Apple has posted an official page with links to download the apps and games on iTunes.
Here’s the tweet from the official App Store Twitter account:
There’s something about seeing the words “innovative” and “landmark” in such close proximity to “free” that really brings home the fact that Apple just isn’t interested in enabling sustainable businesses on the app store. This celebration is not about developers; it’s about users getting treats for having adopted the platform.
To be clear, Apple has been public about this for a long time. David Barnard sent me this screen capture from the 2010 WWDC keynote when Steve Jobs introduced iAd:
The pool of time users spend on smartphones is staggering and growing rapidly, but it is not infinite. The more time people spend with useful/entertaining free apps, the less need they have to actually pay for apps. That doesn’t mean people will never pay for apps — the market for paid apps has continued to grow alongside free and freemium apps — but users have been conditioned to expect more and more for less and less.
It seems they’ve been conditioned correctly; after all, every user just got “five innovative apps and five landmark games for free.”
I can’t be the only one struck by the commonalities between building sustainable businesses on the app store and NSA spying. Or my moving to Taiwan.
(Forgive me one more meta-ish post; I’m still a bit jet-lagged).
First the app store: the inability to monetize existing customers is a primary reason why it’s difficult to build a sustainable business on apps alone. But there’s a second, more fundamental issue at play as well: the absence of friction in searching for, purchasing, downloading, and installing apps.
On one hand, this is a great thing, especially for users. Finding and installing apps is trivial, easily accomplished on the bus, on the couch, or on the can. And it’s great for developers, as a set; the ease with which apps are installed via app stores likely means exponentially more apps have been installed in the last five years than in the thirty-five years that preceded the App Store.
But, for individual developers, the benefits are much less clear cut. The friction of packaging an app, negotiating retail placement, and distribution in the pre-App Store days meant many fewer users ultimately purchased apps (the web solved the distribution challenge, but none of the marketing and installation ones). However, if you did build one of the apps that were widely available for purchase, you also had many fewer competitors.
App stores have massively reduced barriers to entry, ultimately making apps less profitable than they were previously.
Before the app store, initial success was hard, but once achieved, a sustainable business almost certainly resulted. On app stores, success is much easier, but sustainability is that much more elusive. Any changes that Apple, Google, or any other platform owner makes to enable more sustainable businesses on the app store will only go so far in alleviating this economic reality.
Friction was the foundation of sustainability, and now friction is gone.
NSA revelations continue to dominate Techmeme, weeks after the initial Guardian story about Verizon collecting records on every call on its network.
Having labored as a police reporter in the days before the Patriot Act, I can assure all there has always been a stage before the wiretap, a preliminary process involving the capture, retention and analysis of raw data. It has been so for decades now in this country. The only thing new here, from a legal standpoint, is the scale on which the FBI and NSA are apparently attempting to cull anti-terrorism leads from that data. But the legal and moral principles? Same old stuff.
Allow for a comparable example, dating to the early 1980s in a place called Baltimore, Maryland.
The example involves pay phones and pagers, and the collection of metadata surrounding calls, but not the calls themselves. To requote Simon:
The only thing new here, from a legal standpoint, is the scale on which the FBI and NSA are apparently attempting to cull anti-terrorism leads from that data.
Let’s say Simon is right, and was universally acknowledged as such; I bet the outrage would persist. The problem is the lack of friction.
In Baltimore, those detectives had to identify the relevant pay phones, install a dialed-number recorder on each pay phone, clone the pagers, and even then they often didn’t know who the drug dealers were.
Things are much easier today; global communications is largely routed through a few key backbone and service providers, many of which are located in the US. It’s arguably easier to collect the call records of everyone on the planet – and identify them – than it was to collect the records and identities of those Baltimore drug dealers.
One could argue that friction was the foundation of our privacy, and now friction is gone.
I am typing this in Taiwan, rushing to finish in time for the West Coast AM Twitter rush. I know that when I press “Post” people from all over the world will read the words I wrote just seconds earlier. Soon, I’ll start work with a company that is entirely virtual, allowing me to live anywhere I want.
The Internet has removed the friction of time and place, and I am benefitting greatly.
Of course it cuts the other way; I grew up in the Midwest, which means this isn’t my first exposure to the intersection of work and geographic independence. Usually it’s the jobs that move, while the workers are left behind.
Friction was the foundation of our job market, and now friction is gone.
This blog will soon return to the more concrete world of high-tech strategy, value chains, and app store economics, but underlying everything is the seismic change that is only just beginning: if there is a single phrase that describes the effect of the Internet, it is the elimination of friction.
With the loss of friction, there is necessarily the loss of everything built on friction, including value, privacy, and livelihoods. And that’s only three examples! The Internet is pulling out the foundations of nearly every institution and social more that our society is built upon.
Count me with those who believe the Internet is on par with the industrial revolution, the full impact of which stretched over centuries. And it wasn’t all good. Like today, the industrial revolution included a period of time that saw many lose their jobs and a massive surge in inequality. It also lifted millions of others out of sustenance farming. Then again, it also propagated slavery, particularly in North America. The industrial revolution led to new monetary systems, and it created robber barons. Modern democracies sprouted from the industrial revolution, and so did fascism and communism. The quality of life of millions and millions was unimaginably improved, and millions and millions died in two unimaginably terrible wars.
Change is guaranteed, but the type of change is not; never is that more true than today. See, friction makes everything harder, both the good we can do, but also the unimaginably terrible. In our zeal to reduce friction and our eagerness to celebrate the good, we ought not lose sight of the potential bad.
We are creating the future, and “better” does not win by default.
To be clear, this blog is not “famed”! Also, stratechery launched on March 25.
It’s interesting how some folks are always looking for some sort of institutional authority.1 I’ve been quoted as “Microsoft’s Ben Thompson,” as “former Apple intern Ben Thompson,” and “batshit crazy Ben Thompson.” I actually wish the third were true, because, unlike the first two, the descriptor rests on what I write, not on some sort of vague authority derived from whoever is signing my paychecks.
Besides, both workplace references are out-of-date: I was at Apple three years ago, and, as of July 1, I don’t work for Microsoft either. Instead, I am the author of Stratechery. What more is there to say? I’m a person, I put myself out there on this blog, and I trust that what I write represents me well.
One of the many transformative aspects of the Internet is how it empowers individuals to build their own institutions. In days gone by, my thoughts would have been confined to myself and a few close friends; now my friends are all over the world, and I communicate with them through an institution of my own making.
And yet, that sentence is oh-so-presumptuous. This blog has benefited hugely from a few key links and endorsements I’ve received to date,2 and on the technical side, the idea that I built Stratechery is ridiculous. Like millions of other writers who have built something on the Internet, the tool that made it possible is WordPress.
And so, as I continue to build out my place on the web on a blog built on WordPress, I couldn’t be more excited about actually working for Automattic and Matt Mullenweg, the creator of WordPress. I start the end of the month.
This is great news for this blog; WordPress’s original purpose was to be an elegant, well-architectured personal publishing system; the idea I speak for Automattic or that I derive any sort of “authority” from my employment is ridiculous, hopefully obviously so.
What that means for Stratechery is my full, unfettered opinions about all companies and all topics.3 I expect my work to stand on its own, and I hope to be known simply as the author of Stratechery.
One final piece of news: I just said I have friends all over the world – those aren’t just Internet friends. I previously lived for six years in Taiwan, my wife is from here, and we’ve decided to move back. In fact, we arrived this morning, July 4. I’m taking advantage of the fact that Automattic is a company that truly lives its commitment to the web: it exists completely on the Internet, and its employees live all over the world.
You may have noticed a slightly fewer number of posts over the last few weeks; blame packing and the logistics of wrapping up one’s life in one country, and preparing for life in another. I expect the slower pace to continue for at least the next couple of weeks.
I’m not picking on the Quora questioner actually; I’m assuming he/she wanted more information about me and my background. I trust the questioner will forgive being used as a rhetorical device ↩
There have been several others, but those two posts produced the greatest number of return readers by a pretty significant margin; I’m very grateful. ↩
The recent focus on apps is an example; I hadn’t written about apps previously because that was my focus area at Microsoft. To be clear, Microsoft’s blogging policy is very permissive and employee-friendly; I largely self-edited because I felt it was the right thing to do ↩
