A Microsoft reorganization does seem to be imminent. I don’t know how much weight to put on the rumors, but if there is any truth to them I think I’m picking up on a secondary theme to the “Devices and Services” pivot of the organizational structure. That’s a rethink of organizing the company around businesses with independent Presidents.
I provided some history of the current organizational model and thinking in a blog entry back on June 3. The key here is that Microsoft’s last major reorganization was intended to create separate business units with leaders who would be given great independence in running those businesses. Basically the President of a business would negotiate a set of goals with Steve and then be largely free to operate the business without central interference. Essentially, Revenue and Contribution Margin commitments. Overall strategy agreement. And then a bunch of other operational metrics (e.g., customer satisfaction) the company was trying to address. But after that, in theory, the President runs the business.
Microsoft’s current organizational model was inspired by organizations such as General Electric. GE’s Housewares and Jet Engine businesses have nothing to do with one another and are different in every way imaginable. Thus pushing as much leadership to the business unit heads as possible makes perfect sense. Berkshire Hathaway functions in even a more pure conglomerate model, with a tiny corporate office that focuses on capital allocation. Warren Buffet doesn’t try to be an operational manager at all. He buys businesses he likes and makes sure they have leaders he trusts. Then lets the leaders run the businesses. The model has obviously worked well for both companies over a long period of time.
The problem for Microsoft is that it is neither a single business nor a conglomerate like GE or Berkshire Hathaway. Microsoft is a collection of mutually reinforcing businesses. Thus creating business units with leaders who were notionally given independence never worked right. On one hand they never had the level of independence that the organizing principle called for. On the other hand, the independence they had severely damaged the reality that they are mutually reinforcing.
Why, up until recently, was the problem of Microsoft products reflecting their organizational structure becoming worse? Well, if you run them as independent businesses then their leaders are tasked with maximizing the success of their business, not the company as a whole.
In terms of independence Microsoft Presidents had a lot, but it never worked out to the level of something like GE. For example, business units don’t have their own sales forces at Microsoft. Each business unit President has to negotiate with Kevin Turner for sales support. At one point STB wanted to get more aggressive in going after high-end Enterprise sales but that particular proposal was derailed by KT. When Bill Gates was CSA he attempted to apply strategic technical coordination across the independent business units. Bill succeeded, to varying degrees, via his moral authority. But shortly after he left the role the institutions he’d created to carry on were eliminated under pressure from the Presidents.
With the advent of Windows 8 Steve started to apply more pressure on the business units to cooperate and they all did work in support of Windows 8. At the same time you can still see the conflicting priorities in what they produced, with Office (for example) having mostly failed to jump on the Metro bandwagon. Subsequently the Senior Leadership Team took on Ben Franklin’s “We must all hang together, or assuredly we shall all hang separately” attitude. So Microsoft’s key organizing principles is now in conflict with the way its senior leadership is trying to operate. And that is something that should be corrected in this reorganization.
There are many hints in the reorganization rumors that Microsoft will indeed eliminate the structure of having independent business units led by Presidents (and they may or may not use that title for the leaders of new units). The rumors suggest that the leaders of the new units will not have their own CFOs. The CFOs, if they even retain that title, will instead report up through the finance organization. Something I speculated on in the June 3rd posting, that the financial reporting structure will change, is also now being rumored. Yet it doesn’t seem to me that the organization structure being suggested matches up with a reporting structure that will make sense to investors. And both the leaders being suggested, and the terms being used to describe those leaders’ roles, do not suggest they meet the description of a President. Hints might also be found in what happened after Steven Sinofsky left Microsoft, with his responsibilities divided between a business executive and an engineering executive who both report to Steve.
What the tea leaves suggest is that Microsoft is going to broadly split engineering and business responsibilities. Most of what we have heard about until now are the names of the engineering executives and their potential new roles. For example, Terry Myerson would become head of Windows and Windows Phone Engineering. With the departure of Don Mattrick, Julie Larson-Green now seems slated to become head of Devices Engineering. There are two clues here. One of course is that if Mattrick was picking up more engineering responsibilities but losing his business ownership that might have made staying at Microsoft less interesting. The other is that while many would question if Terry and Julie are ready to be President of such large business ventures, their readiness to run the engineering portion of those areas is far more proven. From a people perspective it’s the right next step for both of them.
Talk about current Online Services President Qi Liu also use the word Engineering in describing his new role. Oddly, only the references to Satya Nadella continue to suggest ownership of a business unit. But that may just be a reporting anomaly.
More recently the rumors have suggested new business roles for Tony Bates and Tami Reller. So we could be looking at a return to Microsoft’s 1990s principles in which engineering units run by Group Vice Presidents reporting to the CEO (or, earlier, Senior Vice Presidents reporting to an Executive VP) had significant but not complete business ownership. Or we could be looking at something even more dramatic, a complete separation of business ownership from engineering. For example, it isn’t clear in the new structure if Product Marketing responsibilities remain with the engineering leaders (as in the 1990s) or if those move into an alternate (and somewhat) parallel structure.
If I had to guess, and it truly is a guess, Steve has decided that the way to put the “mutually reinforcing” notion back at the core of Microsoft is by centralizing the business responsibilities to an extent never before (or at least not in the last 20 years) tried. I have mixed feeling about this, which I’ll write about should it come to pass. But while it’s fun to watch my former colleagues’, and in some cases friends’, names being thrown around in new (apparently engineering) roles the real action is in the areas that have so far gone underreported.
Who are Microsoft’s new business leaders and what is the new business structure?
I know Myerson is well thought of internally, but he seems like an odd choice to go to Windows given the criticism of WP’s development pace under his leadership. And Larson-Green moving to devices seems odd given that she knows Windows cold, and also managed to get 8.1 completed on a semi-reasonable schedule. Agree that investors aren’t going to like anything which attempts to reduce operating transparency. I can think of some legitimate reasons to do that, but others could see this as yet another attempt to bury the bad news (like Bing’s ongoing losses, etc) in the good. Of course IF this new structure results in a more agile and competitive MS and net profit/EPS increase, all will be forgiven. What I don’t like is how much of a soap opera this reorg has turned into with daily leaks that appear to be more and more accurate and highly placed. It makes MS look like they can’t keep even the most strategic things secret.
Actually Windows Phone’s development pace has been nothing short of amazing. They did three releases in the time it took Windows to do one and have been on an annual cycle for a few years now. Most of Windows Phone’s problems relate to being late to market (not Terry’s doing, he was brought in to fix that after the fact) and to a strategy that over-focused on the iPhone but reached the market after Android had become the real challenger for OEM and Carrier mindshare. Terry does share a significant part of the responsibility for the over-focus on the iPhone. But it’s not like anyone in the management chain up through and including Steve and Bill disagreed with that strategy. It was what they brought him in to do.
Lots of people know Windows cold. Devices is a new area that provides a good challenge to any existing Microsoft exec, so why not give Julie a shot at it? This is as much, or more, a question of leadership ability than it is being an expert in a particular subject area. If Julie is a great leader she should be able to turn Microsoft into a great devices company.
One more thing. I don’t think Microsoft will eliminate the business segment reporting, just change it. For example, the lack of transparency on the Devices business is bothering investors. And in the near future Devices will become far more material to Microsoft’s earnings results than Bing (which is close to breaking even). So if you lost Bing-specific results in a larger Consumer Apps and Services business segment but gained visibility into the Devices business it would be a big win for investors. The trick for Microsoft is that a reporting segment that combined Online Services including Bing, Live, and Office Client is going to be expected to achieve Office-like margins. And to the extent Microsoft can pull that off over the course of a couple of years, who gives a damn about the specific spend rate on Bing? Similarly, combining Windows and Windows Phone into an OS reporting segment that doesn’t include Live would give far better visibility into this critical business than the current segment reporting. And all this assumes that the segment reporting has anything to do with the engineering organization structure.
Microsoft needs to have a reporting structure that is useful for investors over the next 5-10 years. That isn’t the one currently in place.