What Microsoft’s reorg says about its future (Hint: It’s not trying to be Apple)


Steve Ballmer took the wraps off a mammoth reorganization of Microsoft this week, outlining a new “One Microsoft” structure (and then outlining it again) based on a small number of functional groups rather than a large number of product groups. The basic idea is simple: get product teams and technologies that depend on each other together so they can work more efficiently towards Microsoft’s vision of being a one “devices and services” company.

Microsoft’s new structure looks like a paean to Apple, which famously reorganized around functional groups rather than products when Steve Jobs returned to the fold in the late 1990s. But Microsoft’s business is fundamentally different from Apple’s – and orders of magnitude larger than Apple was when it re-organized at the nadir of its fortunes. Will Microsoft’s new structure produce a nimbler company that can lead technology trends rather than follow them? Can “One Microsoft” be able to make cool stuff people actually want? Or – with almost 100,000 employees worldwide – is Microsoft’s business simply too big to turn around?

Who cares how Microsoft is organized, anyway?

We know: corporate organization is pretty boring stuff. If you’ve ever worked a company job, your first on-the-clock nap was probably at a mandatory meeting with executives monotonously clicking through slides filled with boxes, arrows, and lines. Instant snooze.

Microsoft-Store-32But Microsoft’s organization has always been weird for a company of its size. Microsoft built itself around single products or suites, with each group running a bit like an independent shop. Microsoft also likes its product teams under-staffed and under-resourced: the idea is that by keeping them a bit lean and mean, they must work efficiently, force through roadblocks, and do great work. (If this sounds familiar, today’s start-up companies often operate on similar principles.) Separately, Microsoft Corp. handled things like human resources, finance, facilities, technical support, and infrastructure for everybody: each product may have been its own shop, but they had the logistical backing of a corporate behemoth.

In some ways, this served Microsoft well: Windows, Office, and Exchange each dominate their industries. Microsoft also successfully built its Xbox business largely by keeping it (and its culture) separate. Further, many a Microsoft employee has found working on smaller products rewarding: they’re on tight, fast-moving, motivated teams doing new work, not cogs in a corporate giant.

One could argue that Microsoft’s strategy is the opposite of Apple’s

So what’s the downside? For one thing, separate product teams tend to form cocoons (or “islands,” as Ballmer described them) that don’t keep up with (or necessarily trust) what other teams are doing. If two (or five) teams run into the same problem, odds are they’re going to solve it two (or five) different ways without talking to each other.

Another danger is fiefdoms. Inevitably, lean-and-mean product teams develop a sense of pride and compete with each other. That can make for better products in the long run, but some products are always more important than others. For instance, Microsoft sunk hundreds of millions into developing tablets a dozen years ago, only to have its Office VP essentially torpedo the idea because he thought styluses were non-starters. That’s why using Office on old-school Windows tablets and convertibles sucked. (But yes: styluses probably had something to with it too.)

A third problem is that, at Microsoft, Windows always wins – and that was especially true under the now-departed Steve Sinofsky, who headed up Windows 8 and Surface. Everyone needed Windows, but Windows didn’t need everyone else, so Sinofsky basically played hardball until he got what he wanted. Remember the never-born Courier tablet? What about Danger and the quickly-killed Kin pseudo-smartphones? Both casualties of going up against what would eventually become Windows Phone and Windows RT. Another clash with Windows was partly responsible for cloud computing lead Ray Ozzie leaving the company, and even current Nokia CEO Stephen Elop left his gig as head of Microsoft’s business software division in part due to conflicts with the Windows juggernaut.

OK, but someone has to be in charge, right?

These conflicts make sense, but they also seem unavoidable. Someone has to be in charge of a product, right?

Not necessarily – and Microsoft’s new structure reflects that. Instead of organizing by product team, Microsoft’s development efforts will fall into four main groups. The idea is to break down Microsoft’s product efforts into functional areas so specific people or teams actually responsible for a given function or feature are the only ones working on it – they know it best! – and major decision-making happens towards top of the company hierarchy. In functional organizations, each group only has people who work in its area. A marketing group would just be marketers, it wouldn’t also have its own separate IT department.

SteveBallmerChiefEx header“Process wise, each major initiative of the company […] will have a team that spans groups to ensure we succeed against our goals,” wrote Ballmer. “Each major initiative will have a champion who will be a direct report to me or one of my direct reports.”

As an example, the team working on the next Surface device will have people from the operating system, devices, services, and cloud groups (at least!), ensuring that not only are all those groups talking to each other about Surface, but that no individual team has a stranglehold on others.

And, yes: I’ve simplified. There are other groups like marketing, research and development, business development and evangelism, finance, legal, and more. There’s even one pure product fiefdom left: Dynamics, basically the business management and customer relationship businesses Microsoft’s acquired over the years.

Isn’t this like Apple?

Critics have noted a functional organization is like Apple. In some ways, that’s disingenuous: the details of Apple’s corporate structure are closely guarded, and even long-time employees don’t know much about how Apple is organized. Steve Jobs famously characterized Apple as “the biggest startup on the planet.”

But there are parallels. For instance, Apple does not have a Macintosh division, an iPad division, or even an iPhone division. Instead, product responsibilities are spread horizontally across several roles, including design (Jonathan Ive), hardware engineering (Dan Riccio), software engineering (Craig Federighi), Internet software (Eddie Cue), and technology (Bob Mansfield). The iPhone and the iPad might be driving the company’s bottom line, but neither of them has an executive-level boss. Apple doesn’t even have a head of devices – an odd situation for a hardware company.

Microsoft is taking a page from this book: most products will be driven by teams that span its major groups. There is no Xbox or entertainment group, there is no Office group – heck, Microsoft isn’t even making much of an organizational distinction between consumer and enterprise. There’s not even a Windows group! Sounds like Apple, right?

Apple doesn’t even have a head of devices – an odd situation for a hardware company.

Well, at Microsoft, Windows always wins – and that’s reflected in Microsoft’s new organization, where operating systems not only includes Windows, Windows Phone, and Xbox but also all the “core” cloud services that tap directly into the operating system. The services and cloud groups have already been cut out of that loop. Similarly, some of Microsoft’s groups aren’t actually functional. Sure, there’s a devices group for hardware – one half of the “devices and services” mission! Presumably they’re working on Surface and other stuff, but they’re also responsible for “studio experiences,” including video, music, and games. Apparently they didn’t have enough to do, or there’s a parallel universe were materials testing and system-on-a-chip design is the same business as character art and game mechanics.

Apple may be one of the most valuable companies on the planet, but functional organization was born when it was teetering on the edge of bankruptcy and was ready to accept tremendous change in order to survive. Steve Jobs returned in 1997 with that kind of change, radically re-aligning Apple while jettisoning just about everything he felt was inessential. And Apple has very few products for a major company.

Today, Microsoft is not in the kind of trouble Apple was facing in 1997. So, its reorganization is nowhere near as radical (nothing has been jettisoned), and its line of products and services is much larger. Microsoft’s reorganization is not designed to save the company: it is designed to make it operate more efficiently and (therefore) earn more profit.

Follow the money

moneyLet’s look at Microsoft’s income. Microsoft’s biggest problem is a shrinking cash cow in desktop operating systems – it’s trying with phones and tablets but so far hasn’t seen much success. (Its Android licensing program, however? Pure gold.) Microsoft also has major income from mainstream productivity software – and it’s gradually converting Office customers over to annual subscribers rather than occasional license-buyers. A third cash cow comes from back-end enterprise and cloud services.

One thing these business have in common is that their most important customers are volume licensers (like computer makers, enterprise, schools, and governments), not individual consumers. Yes, Microsoft has other businesses, and some are even consumer-focused. Microsoft runs the number-two Internet search engine, has its own gaming empire (Halo, anyone?), and runs some of the world’s most-used Internet services like Outlook.com and Skype, and has its own games/music/video/software store. But right now, those things aren’t major contributors to Microsoft’s bottom line. (In fact, collectively, they’re a drain.)

Microsoft’s biggest money comes from its biggest customers, and Microsoft’s reorganization is mainly about keeping those customers on board and happy. This is plainly evident in the emphasis Ballmer places on applying technology to meetings and decision-making, describing them as “time-intensive, high-value scenarios that are ripe for digital re-imagination.” To do that, Ballmer calls out Bing, Excel, and InfoNav (an in-development natural language query engine – think Siri) as key technologies to help with decisions and tasks. Ballmer says “we have focused not only on what matters most to individuals but also on what’s vital to businesses around the globe.” But right now, the corporate customers are the ones who are buying. Outside of Xbox, Microsoft’s success with consumers has been very limited.

Will Microsoft get cooler?

Microsoft has famously missed the boat on key technology developments of the last two decades. It didn’t anticipate the rise of the Internet; it was very late to personal media players (remember Zune?); failed several times in digital media; got to the Internet search game late; whiffed on blogging, social networking, and sharing; took a $6.2 billion writedown on its online marketing business; and it’s now very late to smartphones and tablets (now discounted!).

This doesn’t mean Microsoft is incapable of producing technological innovations that will change the industry. However, despite breaking down most of the formal lines separating its consumer and enterprise businesses, Microsoft’s eyes are still going to be on its own bottom line. For now, that means enterprise, and the company’s “one Microsoft” vision is mainly about making it easy for those enterprise customers to continue using Microsoft technology, in Ballmer’s words, “after hours.”

In fact, one could argue Microsoft’s strategy is the opposite of Apple’s. Apple’s recent successes in the workplace and enterprise has largely come from customers using their iPhones and iPads to work whether employers liked it or not. Microsoft is effectively trying the reverse: hoping enterprise users will take Microsoft products home with them. For some customers – particularly those who live in meetings and are “tasked” with “decision-making” – that might be a no-brainer. For others, it’ll be an uphill battle.

Get our Top Stories delivered to your inbox: