Steve Ballmer announcing his intent to leave Microsoft was unexpected and potentially a watershed moment for the company and also for the industry. The shift of a major tech company from founder-class management is a big deal any day, and it is a bigger deal when the company is of the scale of an $80 billion run-rate. Yes, Ballmer is not technically a founder, but as employee number 30, brought in by his founder best-friend at a very early stage, it is a technical distinction.
While Microsoft’s stature and overall industry relevance have been eroded by the shift to mobile, it is unfair to argue that Ballmer’s tenure has not been a success, at least by the numbers. An analysis using data from the Fortune 500 public website shows that, during the period he ran Microsoft, Ballmer is third only to Exxon’s Rex Tillerson and GE’s Jeff Immelt in the dollar-volume of company profits he has delivered.
But profits are not everything and many of Microsoft’s problems, like those of most for-profit businesses, can be attributed to decisions made to maximize profit at the expense of customer loyalty or adoration — a quality Microsoft now finds itself craving and coveting as it has entered the devices space in competition with its arch-rival Apple.
The Ballmer CEO era should be understood as a natural extension of the Bill Gates era and indeed it is hard to disentangle the impact of technology decisions made by the two CEOs. The famous delays in Vista — which Ballmer claims as his biggest failure — are fundamentally the result of decisions around Longhorn which represent the company’s grandest failed over-reach. Longhorn was decidedly Bill Gates’s technical vision, who was the Chief Software Architect at the time. Business books will analyze the Microsoft “founder era” for years to come, but to my thinking, the company has been caught in a classic Innovator’s Dilemma, where its success during one era precluded decisions that were essential for its success in a new era. In other words, there is an element of inevitability to some of its failures as no company can be at the top of the market forever.
This does not mean the company cannot make course changes and shift the trajectory of its relevance in the industry, and indeed potentially become the top mobile device maker in the world. In fact, Microsoft has made some significant changes over the last four years, setting itself on a trajectory of success if it can execute and adjust rapidly enough as it goes. In the spirit of looking mostly forward, while glancing occasionally at the rear-view mirror, I propose ten technology decisions that Microsoft’s next leader should consider.
1. Ship Office for iPad and Android tablets
This is a decision that has to be made soon. The Office franchise is exposed to erosion as non-Windows mobile devices take productivity work away from Windows PCs and mobile devices. Microsoft has always thought of integration of applications and platforms as a sacred right and fought for it hard in the DOJ case, but it has to rethink its vision of the dominance of Windows in the mobile world and adjust to the reality that it will be at best one of several major players in future devices and will never have the control it did in the pre-touch era. Supporting non-Windows tablets full bore will maximize Microsoft’s paths of eventual success.
2. Pick either Windows Phone 8 or Windows RT to fight the mobile ARM-chip OS battle
Microsoft is in a strange situation with its mobile strategy as this big proponent of focus, synergy and efficiency finds itself with two different operating systems in the mobile space being pitted against iOS and Android. True they now have a converged kernel, but they have two different application models and user interaction models that are not insignificant to converge. The Windows Phone team moved faster than the Windows team to adjust to a new mobile world, and that explains the situation to some degree, but Microsoft should have taken the opportunity with the Windows Phone 7.5 to Windows 8 transition to align with Windows RT. It is better late than never, and at the expense of another ecosystem disrupting transition, this has to be done soon.
3. Invest in evolving the Windows 8 Desktop application model to recognize that sit-down personal computing remains a sizeable and profitable long-term franchise
There is no doubt that future growth in devices will be around phone and tablet touch form-factors. It can be argued that PC sales of the last ten years were heavily bolstered by lack of alternative form-factors to the traditional clamshell PC for basic digital connectivity (e.g. Skype, YouTube, Photography, etc.) Today these workloads are dominated by $150 to $300 touch devices. Nevertheless, a sizeable market remains around sit-down professional computing which puts a premium on larger screens and a console-like productivity experience. This is basically what the Windows Desktop was designed to do and does well.
Microsoft’s recent focus on touch has come at the expense of too much doubt that the Desktop is about to go away, sending shudders of anxiety in its traditional user base and paralyzing them from potentially investing in replacing their PCs. More investment in desktop will provide a base of support for higher-priced PCs, which can be a sizeable and profitable market for a long time to come, even as cheaper devices dominate the growth of casual computing.
4. Ship a $99 ‘Xbox Light’ for entertainment, mobile casting and casual living-room gaming
Until recently, Microsoft’s experiment with Xbox could be seen as one of its best turn-around stories and a genuine success. However, traditional console gaming is giving way to casual gaming, and devices in the living room, including smarter TVs, are proliferating and growing. The current plan to compete at $499 will all but stifle any possibility of a mass market materializing for casual gaming built on Xbox technology. Microsoft needs to find ways of integrating its mobile technology and ecosystem with a broader presence in the living room. This can only be achieved by playing with the price elasticity curve.
“Devices and Services” sounds great on paper, but trying to follow Apple with a head-on confrontation may be too late and can hurt ($900 million of hurt in write-offs for unsold Surface RT machines). Just as importantly, there is no easy way to negotiate the OEM ecosystem, which was instrumental to Microsoft’s PC success, by competing with it half-heartedly. Most importantly, Microsoft needs to build credibility and know-how in devices slowly and deliberately. For this, a strategy of serially partnering to use its highly promoted Surface brand with other manufacturers may speed ideas to market and defray the self-hurt and the ecosystem hurt.
6. Provide a Bing search engine which does not store personally traceable data
Microsoft has tried everything in Search and may be close to making some hard and difficult decisions on future investments. Bing is a solid alternative to Google Search but continues to play catch-up to Google’s well-monetized R&D agenda in the space. Out-of-the-box thinking is required here, and we are seeing signs of it with Bing by noting the recent offering to schools of advertising-free search. Given the developing anxiety users are feeling these days about surveillance, Microsoft should consider upping the ante and offering data-retention-free search services to consumers that are difficult for governmental bodies to surveil. This may compromise certain lines of advertising products, but might be a lead service to bring in users who might then opt to share more for added services.
7. Move R&D in the company to a 12-month major release cycle and 6-month or 3-month minor release cycles
Changing how software R&D is done is hard and Microsoft indeed has begun to tackle this. Shipping Windows 8.1 with some significant changes in a one year cycle is an example. The reality is that software evolution needs to move faster still if the company is to catch up or keep up with its competitors. Big-wave software releases with multi-year R&D cycles are not just a sign of self-indulgent development processes, they are a symptom of software whose primary competitors are prior releases of the same software. In the more competitive world Microsoft finds itself in today, the harsh disciple of schedule-driven software update cycles, including for major new features, is essential for success. If there is not enough time to introduce innovation in this time-frame, a parallel team should to be dispatched to incur whatever time and cost needed to re-architect the software so that appropriate evolution can fit into the release-cycle schedule. The era of devices and services in today’s high-velocity business environment demands nothing less.
8. Embrace open source software (OSS) for big products, starting with key developer technologies such as Visual Studio.
Microsoft has been fearful of IP pollution and other considerations of OSS even as it has dabbled with it byreleasing small projects here and there. In fact, the extensive use of OSS inside of Microsoft products could help move the technology cycle-time faster. In various flavors and degrees, OSS has been proven as a business model and as an approach for continuously managing some of the largest software projects in the world (e.g. Linux). Microsoft’s battles with OSS over the years have precipitated considerable developer aversion and even hate which has not been helpful in the broader adoption of Microsoft technologies. Just as seriously, it has isolated Microsoft from great talent in the industry and stymied its ability to see new trends in the market and interpret them because of lack of such talent. Fostering the love of developers, especially those of other platform ecosystems, may be one of Microsoft’s key tools for future growth in relevance and consequently its own platform’s adoption. As open source software continues to grow, Microsoft’s leaders and legal team have to come to terms with better leveraging it to shift Microsoft’s image and its agility and effectiveness to compete in the new world. Moving technologies such as the IDE to an OSS model may be just the right injection of love into its developer ecosystem.
While similar in impact to broader OSS adoption, embracing Linux can potentially bring more than goodwill. When Windows NT was being put together in the 1990’s, server computing was dominated by expensive, big-iron Unix machines. Microsoft correctly saw the trajectory of x86 servers as the slam-dunk to future dominance of server computing with Windows NT. But Linux happened, and Microsoft spent a great deal of time and money, and executive hair-line, on fighting it in high profile marketing battles. Linux did not go away. While Microsoft has achieved a great deal of success on the server side, Linux dominates web workloads and remains a chronic threat to be managed for Microsoft. Linux powers large Internet companies and some of Microsoft’s most formidable competitors like Amazon and Google. Microsoft’s aversion to Linux has not only cost it enterprise adoption of technologies like SQL Server, but also complicated the infrastructure strategy and human-resource aspects of acquiring companies like Yahoo!. To be sure this is something that appears to be changing as Azure has embraced Linux distributions already. In the age of cloud computing, the SLA (Service-Level-Agreement) is an equalizer of underlying technologies. Still, supporting Linux will avail Microsoft of a considerable body of know-how in the industry and drive the company to work together with the industry, and not against it. Embracing Linux more whole-heartedly by porting other server applications to it might translate into more enterprise adoption of such offerings.
… and, the 10th technology decision a new Microsoft leader should consider
10. Bring back the traditional desktop Start menu so that Windows 8 feels like a comfortable transition from Windows 7
On an R&D complexity scale, this is a trivial move that can yield a great deal of Windows 8 new adoption bang for relatively little bucks. Aside from the many users who are befuddled by the lack of the original Start Menu, there is the problem of PC OEMs, such as most recently Lenovo, who are by-passing Microsoft’s user-experience vision by including a Start Menu substitute.
Many years of market dominance have created a sense of entitlement inside Microsoft, leading it to occasionally follow approaches to force adoption of new capabilities by removing the old ones cold-turkey. While this may be effective for a dominant player, this approach can backfire in a competitive market when there are alternatives. With Windows 8.1, Microsoft brings back a Start Button which will continue to throw the user in the tile-mode of Windows, requiring the learning a trick or two to get back out into the traditional Desktop. A more competitive Microsoft would view user comfort and stability of user experience across software releases as a marketing and positioning value-proposition to be leveraged, instead of a strategic obstacle to the direction of its vision.
Al Hilwa is an industry analyst at research firm IDC specializing in application development research. Mr. Hilwa has written columns for various tech publications and is widely quoted in the media, including the New York Times, Financial Times, Wall Street Journal, and National Public Radio. Mr. Hilwa joined IDC in 2007 after seven years at Microsoft where he held various positions as product manager and strategist in the Server & Tools division. Prior to Microsoft, Mr. Hilwa was an industry analyst at Gartner and, prior to that worked in various IT roles in various industries. Mr. Hilwa holds an MS degree in Computer Science and a BA degree in Mathematics and Computer Studies.