Hardcore Software has shared the vision planning process for five releases of Office and Windows 7. Though not detailed we followed the same process for two waves of Windows Live Services as well as Internet Explorer 8 and 9. Windows 8 went through this same process, though by now as a team we had become pretty good at it. This section details the resulting Windows 8 plan, The Vision for Windows 8. As part of that, I wanted to take a bit of a journey into the alignment between Windows Phone and Windows 8 and the challenges we saw there. In doing so, I will describe things from the Windows perspective and not delve into the specifics of running the Windows Phone project, which wasn’t my responsibility. Rather, I wanted to cover the challenges of two large projects within the context of Microsoft each trying to figure out what they needed to do. Since 2010-2011 when this took place, it is the strength of Apple’s approach of starting from a reinvented desktop operating system for the iPhone and building out from there that makes the events of this time strategically interesting. As I frame events, the key questions to ask would be “Should Microsoft have waited?” and “Would it have been ok to not be in market with any phone after Windows Mobile 6.5 until 2012 or even 2013?” At least that’s how I reflect on these times. The answers are not complete as the next chapter will also cover some important aspects of this in more detail, particularly the hardware and platform elements.

This section could really be a chapter and isn’t for the faint of heart. Dig in and have fun because it covers a lot of ground that took place in a relatively short time.

Back to 099. The Magical iPad

We never lacked clarity in what we intended to do—reinvent Windows for a new era. That meant a new experience, a new platform for developers, a new connection to services, and, yes, new hardware. Books about reinvention (or disruption) don’t tell you that you can’t just announce such intentions to the world. It turns there is the intention to reinvent and then a plan to build it, though when you decide to share a strategy is an entirely different matter.

I was old enough to have personally lived through the creation of the term Osborne effect1 as it pertains to pre-announcing products. In high school I programmed my father’s Osborne he bought to keep the books for the family business. We’d been using the Osborne daily since it launched and thought about buying an upgrade for the business. Then Osborne founder and CEO, Adam Osborne, pre-announced the Executive, a fancier model with more memory and a bigger screen. The only problem was it was far from being done. The prototype product was shown in 1983. Customers held off buying the original Osborne long enough that the company went bankrupt before delivering the new computer.

Apple faced a similar dilemma when it first transitioned from Motorola chips to PowerPC chips. The transition was announced in 1994. While it is difficult to tease out the impact of Windows 95 from the chip transition, Apple’s share of the Mac/PC market would steadily shrink for another decade, and decline significantly in absolute unit sales, until the next chip transition to Intel.

Microsoft had long been relatively immune from pre-announcing products because the overall growth of the PC market combined with the breadth of the product line dampened any pullback in a single product. The PC needed an OS even if the next one was delayed. As we saw with Longhorn/Vista, businesses still continued to buy PCs in droves. That’s why many Microsoft products seemed to be talked about long before they were released. There was an added benefit to this early sharing, or as we called it openness, which was used to generate platform momentum. With nothing to lose, Windows itself benefitted from a solid 5 years of momentum building in the early days before Windows 3.0. Back then it was all just an industry norm.

In our world, the Windows business had just survived the Longhorn mess and recovered with Windows 7. We now faced an entirely new market situation. Windows itself faced structural challenges—actual alternatives in the market in the form of phones, tablets, browsers, Intel-based Macs, and soon ChromeOS. At the very least, people could just stick with Windows 7, which was fine by us, except given new alternatives most customers would not even consider buying a new or additional PC, which was very bad for us.

Every bone in the Microsoft body would cry out to begin evangelizing Windows 8 as soon as we had plans. We were going to build a new platform and evangelists wanted time to articulate the strategy so developers could weigh their alternatives. But doing so would also run up against Windows Phone 7 and the platform they were evangelizing. The lack of strategic connection between Windows and Windows Phone was obvious but at the time was fraught with difficult choices, especially in the context of competing with Apple.

Then there was the biggest of all problems, again something they fail to mention in books. What if the big strategic bet we planned on making ran right up against our biggest partners and customers? The whole idea of advancing Windows without our partner Intel and the major PC makers Dell, HP, and Lenovo would be heresy, plain and simple. We are talking about Wintel after all. Not only were we planning on a chipset, but we knew we would offer something radical with respect to the actual computer we’d offer customers. Double heresy.

As a result, the vision we created for Windows 8 was not specific in the broad communication about the role that alternate chip platforms, SoC or system-on-a-chip, or new hardware would play in the plans. By using the term SoC we could account for both ARM, the then UK company that designed the chips used in all mobile phones and tablets, and Intel who continued to work to develop a competitive SoC with their ATOM branded chips. While the engineering was well underway, the degree of the bet still needed a bit more data and experience to decide if we could execute specifically on bringing Windows to ARM. By using SoC, if the term leaked, we could always point to Intel’s latest ATOM chips as the goal. The cost of openly defying our own ecosystem and then failing to materialize would have been immense given the state of the PC market.

As far as how these alternate chips would come to market, we had not yet decided on a complete plan. Would we go the standard route which was to evangelize to the OEMs as we did with Media Center PC and later multitouch support? Would we build a first-party demonstration device and use that as the basis of evangelism as we did with Tablet PC? Or would we commit to what was either unimaginable or incredibly dumb depending on if you were our customers/partners or BillG and design, build, and sell, our own new device? The vision would be silent on this. The plans were still being made and would be resolved just a few weeks after the product vision was communicated.

With the iPad announcement described in the previous section, the pressure on everyone to respond was immense. For some the rise of Android was even more worrisome, primarily because of the history of Apple being less of a real threat. Google caused more consternation because it was growing so large so quickly in an area entirely new to Microsoft. Either way, even though Microsoft was the among the first to enter the smartphone market, by 2010 our share remained in low single-digits and would never get much higher.

The app revolution underway in Silicon Valley was on the iPhone. And we were missing it. I learned this firsthand sitting in the pouring rain after a wedding in San Francisco in 2010, when everyone else was using a new iPhone app to summon a limousine late at night. It was my first experience with the UberCab on-demand car service, and we couldn’t summon one on my Microsoft phone as we became drenched.2 The app gap was just getting started. Every day it seemed like a new company released a new app for the new iPhone platform and none of that innovation was happening on Windows. I swear it felt a bit like what IBM must have experienced when everything new was on Windows and all that was on OS/2 was what they paid to have there.

Windows and Windows Phone: Alignment?

The Windows Phone Team was in the midst of the significant reinvention of the phone platform, originally code-named Photon, which became Windows Phone 7 or WP7. The fall 2010 release was more than six months from the rollout of the vision for Windows 8.

There were many challenges in entering, or re-entering, the smartphone business. Bootstrapping an ecosystem was chief among them. Finding the right hardware partners proved difficult in the face of competition for those same partners from Android. There was also the challenge of building a differentiated product when the high-end was so solidly iPhone while Android seemed to cover the breadth of the market. To many reading this, the market might appear analogous to the evolution of the PC market, except Android has the role of Windows and the premium niche occupied by iPhone is much larger than the eight share points Apple computers achieved.

WP7 faced several software challenges simply because the core operating system was so old. Support for the latest capabilities across graphics, networking, multicore CPUs, removable storage, and devices (such as the NFC reader required on Japanese mass transit) was becoming increasingly difficult to impossible, as was enabling the product to work worldwide across languages with complex characters and input methods. A hot topic on the heels of the product release was the rollout of 4G or LTE technology, especially in Asia, and the challenges WP7 had in building support in the operating system.

When it came to synergy with Windows the developer platform was the key challenge. Windows Mobile, Microsoft’s original phone OS, supported a subset of the Developer Tools strategy which had a nominal relationship to Windows itself. Breaking with that platform, WP7 made a new bet on the nascent Silverlight project, a decidedly non-obvious strategy relying on a technology project essentially salvaged from the prior big bets in Longhorn as I described when I supported moving the team and early project out of Windows. As a competitor to Adobe Flash, it seemed most suited to lightweight games, video, and photos designed for a cross-platform experience. A significant class of apps, sure, but as a general-purpose platform, a Windows platform, Silverlight was inadequate and would face challenges in maturing to a complete platform. For example, there was little hope of building an email client or Office in any enduring way using Silverlight’s capabilities as they were. That was certainly my opinion. They made a strategy that was first and foremost based on time to market and that leveraged a new initiative in Developer Tools that many were very excited about, including the evangelists. The use of Silverlight felt leveraged and very…strategic. This will be further discussed in the next chapter. Given the break from Windows Mobile, there was an opportunity for a strategic platform alignment that was not taken advantage of.