The web is already full of experienced Microsoft watchers’ opinions. Some of them, like IDC analyst Al Hilwa’s dive into how Microsoft needs to refine its product line, are worth a read. But most miss the more fundamental problem, which is how Microsoft’s failures are a direct expression of its broken management culture. Fortunately, veterans of the firm, speaking on condition of anonymity, were as happy to discuss how to fix that culture as they were to talk about how it got broken in the first place.
1. Be ready to annoy some people
Like the city of Detroit, Microsoft has obligations it’s probably better off without. Bringing in a new CEO is the perfect time for Microsoft to do an accounting of its relationships and figure out whom it needs to alienate in order to refine its strategy—whether developers, resellers, support vendors, or corporations with which it has alliances.
2. Unlock the talent that remains
Despite hemorrhaging senior engineers and executives to competitors, Microsoft is still fairly decent at attracting talent, say my sources. Committed engineers will tolerate a certain amount of friction with management, but the company’s “stack ranking” method, in which the lowest-performing engineer on any given team gets a bad performance review, has to stop. (Sources say this destructive management practice, first outlined by Vanity Fair in 2012, is ongoing.)
Performance reviews are common at big firms, but when they’re aimed at figuring out who isn’t cutting it, they automatically alienate people. One thing Google does much better than Microsoft is to focus on improving employees rather than punishing them.
3. Make the most of an enormous trove of intellectual property
You don’t hear about it much outside the whizz-bang portion of the the technology press corps, but Microsoft Research, the blue-sky research division, is a modern day Bell Labs. Divisions within Microsoft Research, which is spread across six countries and three continents, work on everything from machine learning to the future of human-computer interaction.
The problem with Microsoft Research—and the same was true of the research labs at companies like Bell, Xerox and Kodak—is that it has become too shielded from the need to deliver products that can contribute to Microsoft’s bottom line. As one source put it, Microsoft Research’s current output isn’t “productizable,” and is too focused on how to accomplish an interesting technology once, in a controlled environment, rather than making it suitable for a billion people.
But that may be changing. Software like GeoFlow adds “business intelligence” to Microsoft Excel. In this case, it’s “3-D geospatial and temporal data visualizations,” but Excel 2013 is full of these capabilities. The technology behind GeoFlow was born as a far less business-relevant project, Microsoft’s WorldWide Telescope for visualizing astronomy data. Clearly, there are ways to take the work that has been done in Microsoft Research for decades and turn it into things Microsoft’s prodigious sales team can stick a price-tag on.
4. Figure out what the “next Xbox” will be
A good rule of thumb at Microsoft is that the less a product needs to be a part of Windows, the more likely it is to succeed. Windows is historically important to Microsoft, but represents only 22% of the company’s revenue. Yet forcing engineers to route their products through the Windows team and management, has slowed innovation to a crawl. “Every little thing you want to write has to build off of Windows or other existing products,” one software engineer told Vanity Fair. That’s one reason why the Xbox, its gaming console, and the Kinect, the motion-sensing controller for it, succeeded—given license to proceed at their own pace, Microsoft’s engineers can still create products people want.
IDC analyst Al Hilwa has suggested that Microsoft’s next Xbox could simply be a $99 “Xbox lite” designed to compete with Google and Apple’s set-top boxes, by offering similar capabilities, such as media streaming and casual games. But whatever it is, whether it’s another go at making tablets after the disastrous Surface RT tablet, a watch, or something else entirely, it’s going to be a high-risk venture. Ballmer was neither the technical nor the product genius Microsoft needed in order to create new categories of products consistently, as Apple, Amazon or Google do. Hopefully the next CEO will be.
5. Continue becoming the Apple of enterprise software
Microsoft is growing fastest in the “enterprise” market of software and services for business, which represented $22.4 billion of Microsoft’s $80 billion in revenue for the most recent fiscal year (pdf). Successes in this area include Azure, Microsoft’s competitor to Amazon’s cloud services, which is already producing $1 billion a year in revenue. And the web-based SharePoint system for managing documents within a company is a $2 billion-a-year business, according to a Microsoft spokesperson.
Microsoft needs to put more effort into the enterprise market, and emphasize both to investors and the public that Microsoft, like IBM, is about solving customer’s problems, not selling specific products. Arguably, the company is already the “Apple of enterprise,” in the sense that it’s creating unique products for enterprises that nobody else is, such as SharePoint. In this area at least it’s unambiguously succeeding.
6. Split into multiple companies
In 2001, Microsoft’s monopolistic practices (bundling its Internet Explorer browser with Windows) prompted US antitrust regulators to propose breaking the company up, though they relented in the end. So it’s ironic that a breakup is now the subject of serious discussion among Microsoft watchers. Only this time, it would be about making Microsoft’s various divisions leaner and more competitive.
This could also allow productive mergers. For example, Microsoft’s mobile division could merge with Finland’s Nokia, which is the biggest manufacturer of smartphones running on Microsoft’s Windows Phone operating system. It could then kill off Nokia’s shrinking and fragmented ”feature” (i.e., non-smart) phone business, and instead focus on the Windows phones, which are actually quite good.
Independent Microsoft fiefdoms would also be less hampered by each other’s interests. For example, there is no version of Microsoft Office for Android-based tablets and the iPad, because that might hurt demand for Microsoft’s own Windows-based tablets. But since Windows has only a tiny share of the tablet market, that basically dooms Office to irrelevance except on personal computers—the market for which is shrinking. So the Office division needs more autonomy just to ensure its own survival.
7. And the one thing Microsoft mustn’t do…
In the past I’ve written that Microsoft should abandon the consumer market, where it’s failing. Many Microsoft investors say the same. But my sources said otherwise. They argue that the dominance of Windows on the computers people use at home has long been Microsoft’s main advantage in convincing executives and IT buyers to use the company’s enterprise systems.
Fortunately, not abandoning the consumer could be as simple as undoing some of the more radical recent changes to Windows that have put customers off, and making it more like its old self, “Start” button and all. Microsoft must also keep working on Windows Phone; it’s a good (if late) product and it can at least take a solid third place behind Apple’s iPhone and the myriad devices running Google’s Android OS.
Ultimately, Microsoft doesn’t need to be “saved” so much as re-configured, so that it can continue to grow rather than coasting on the fact that so many companies still use old Windows-based systems. Finding a CEO who can rebuild the company won’t be easy, especially because the one C-level executive who has most recently accomplished a successful re-think of his giant tech company’s core strategy probably doesn’t want the job.