064. The Start of Office v. NetDocs

Microsoft went through so much in the first year of the millennium. It began with SteveB taking on the role of CEO and BillG taking on a new role as Chief Software Architect. Over the first months of the year a new set of technical leaders convened under the direction of Paul Maritz leading all of the product groups to define essentially the next Windows. The group was producing the plans for NGWS, next generation Windows services as outlined in memos from BillG and SteveB. The DOJ trial was complete, and we awaited the verdict, but everything we did was looked at through the lens of the implications of the trial. At every step we were asked if what was going on was a result of the trial, anticipating an outcome, or designed to work around what comes next. Personally, I was just getting my footing as an executive and the leader of Office and just figuring out what it means to be leading such a big part of Microsoft. I still had so much to figure out and was definitely worried about leading the train off the tracks.

NOTE: I really want to offer a big “Thank You” to all the paid subscribers. It takes a lot to take the time to support the work materially and while the proceeds of this work go to registered non-profits in the US, it still means a great deal that you’re paying for the work. Many subscribers are about to hit their one-year renewal and I wish to thank you in advance (and welcome new subscribers). As planned, there is almost exactly another year of posts planned—and for many this is the extra-exciting stuff including what’s coming with SharePoint, Outlook, the Ribbon, Windows 7, Windows 8, Surface, and many controversial (at the time) topics like Courier, including today’s section on NetDocs. THANK YOU!

To celebrate the anniversary, please consider sharing this link for a discount on the yearly subscription. https://hardcoresoftware.learningbyshipping.com/anniversary

Back to 062. Antitrust: Split up Microsoft and 063. Managing a Verdict

The demonstration within the multi-hour series of keynotes was billed as a “sneak preview of something that hadn’t been demoed before. . .technology that embodied the dot net user experience. . .this is real code. . .this technology will apply very broadly in the future across Microsoft products such as Windows.NET [pronounced Windows dot net], Office.NET as well as the consumer subscription service.” What followed was a 15-minute demonstration of word processing, spreadsheet, email, calendaring that looked like Office. The demonstration was easier to use, sleeker, and more connected. It featured enthralling technologies like “universal canvas” and “XML”. What’s not to like?

Within a news cycle the technology demonstration had ballooned to Office.NET and was the future of Office. Within Microsoft, especially in Systems, nothing was higher praise than being the future project and conversely nothing was worse than being the past. The world of dueling code names had been brought to Office, except now it was Office.NET and whatever I was working on, aka Office10. What had been shown was being built by a separate team, an organizational peer of the newly christened old Office.

What was this and where did it come from? Was anyone building a product called Office.NET? Was this planned? I certainly knew the code being demonstrated, but the idea that it was presumed to be the future of Office was newsworthy, even though we did not say that directly and had not intended to leave that impression as far as I knew. Nobody wants to Osborne the most profitable part of Microsoft (a reference to a well-known microcomputer company that went bankrupt preannouncing a next generation product). Weird.

Starting in early January of 2000 coinciding with SteveB’s promotion to CEO and BillG assuming his new role of Chief Software Architect, BillG and PaulMa began working on a series of strategy offsites, meetings, brainstorms, memos, and more called Next Generation Windows Services, or NGWS. There was even a new leadership team formed called the TLT, Technology Leadership Team. Everything was kicked off with memos from both BillG and SteveB highlighting NGWS. BillG explained in a memo Opportunities in the “Software Decade” that NGWS was a bet on par with the graphical interface in both the transformation and opportunity. By now, I was seeing this as a familiar playbook. If you want to say something is big, then compare it to the graphical interface. SteveB also had a memo, Changes and Opportunities.

These memos, Bill’s detailing the technology at a very high level and Steve’s articulating a customer and business focus put forward an innovation agenda for the company. The goal was to let Microsofties know the company was committed to innovation, especially with the rise of “dot com” companies and the huge valuation of anything internet in the public markets. It would be a few months until the market corrected itself, but Microsoft at this juncture was generally in a defensive posture. A broad re-branding was intended to create a new narrative for the company, supported by a next generation of technologies designed for the internet from the ground up, at least starting from Windows.

Just days after Bill’s memo, there was a call for participation in the NGWS planning sessions. A detailed series of offsites and meetings were scheduled. Literally, the invitation stated the work was to figure out the details of NGWS as introduced in the memos.

Please consider subscribing. New subscribers can use this link for an anniversary promotion discount.

In other words, we had a brand before we knew its meaning. That is not entirely fair because there were at least two key strategies under development. A series of projects defining consumer internet services such as email, calendaring, and identity were being developed by a group of the most senior leaders previously of Windows NT and others from around the company, an effort that would later become known as Hailstorm—consumer-focused experiences that also provide a platform for developers. A second effort encompassed the creation of a new programming platform for building internet applications on the server, which was just becoming known as .NET (dot net). This initiative included a variety of tools and platform work that built on the lessons from the first generation of internet applications.

To confuse matters, the .NET branding was starting to be picked up by a range of groups and products, including the next release of Windows NT Server which was sometimes referred to as Windows.NET (originally codenamed Whistler, then later Windows Server 2002, then finally Windows Server 2003). The .NET name was also used for the APIs and programming tools which would collectively be called the .NET Framework, for programming both servers and desktops. If you’re confused reading this, then you were not alone. NGWS was an umbrella term for everything, and .NET was intended to be a technology term, but we sort of ended up with two umbrella terms. The .NET branding was one of the more chaotic and self-inflicted product naming efforts (Was it .net, .NET, or .Net? Before or after the product name? With a space or without?).

Like BillG’s previous memo on software as a service, this memo also lacked any mention of Office. I was beginning to see a pattern. Only now that I was managing Office I started to wonder if I was somehow contributing to this. Whenever I reviewed drafts of these memos, I did not seek to include Office out of a reflex to fly under the radar and to avoid making promises for work that was not even underway, a strategy reinforced by my time working as BillG’s technical assistant. The team always noticed, and I found myself doing my best to explain the virtues of being left out of the strategic fray. Should I have lobbied or been more forceful about including Office? Many would be naturally inclined to run towards the limelight, but so far in Microsoft’s history that proved to be less than helpful. The Cairo OS project was a top-of-mind example.

PaulMa and the platform teams planned a strategy presentation for the press and industry analysts detailing Microsoft’s internet-centric developer strategy. Originally scheduled for early June 2000, the event was delayed several times because of the looming court ruling in the antitrust trial. The NGWS working groups welcomed the extra time. The evangelism efforts did not slow, however, and the first half of 2000 was a steady stream of stories about what NGWS might be along with the implications of the looming court ruling on NGWS, or perhaps speculation that NGWS was an effort to end-run the potential court ruling.

The Windows team generally loved these strategy days as a key part of the culture. PaulMa would often describe them to me as a “forcing function,” which meant a way to coalesce disparate groups into a shared plan. In this case, the planned event would force upon us a collective definition of NGWS.

The industry loved these events too. These were made-for-press events. Stories would run describing what could be announced in the lead-up (called curtain-raisers) and after the event there would be ample analysis. As was almost always the case, the event would gain a nickname or acronym. The event almost always included a new strategy with its own name or acronym. On the heels of Internet Strategy Day and Windows DNA (Distributed internet Architecture), the press was anxious to learn what else was on the way. Frequently, such event days would get scheduled with only a vague idea of what would get talked about and shown. This was one of those events.

The weeks leading up to the event were chaotic and high stress. While the goal was to present a coherent strategy, the process of creating the strategy was more important—the forcing function. This was how Platforms came together as a team. Whether a PDC (Professional Developer Conference), a workshop, or a strategy event, Platforms used the process of creating the presentations the same way Office used memos and the vision planning process. Only Office spent months and involved a broad cross-section of the team across disciplines whereas Windows spent weeks and usually involved the key people, however that might be defined. Instead of detailed plans and schedules like we created in Office, the output consisted primarily of bullet points and architectural diagrams in PowerPoint.

Dubbed Forum 2000, the event proved a seminal moment in the evolution of Microsoft’s platform and quickly came to be known as .NET Strategy Day or .NET Day, and SteveB would refer to it as “most ambitious undertaking since Internet Strategy Day in 1995”. The event aimed to be almost a “mother of all demos”, in reference to the legendary 1968 demonstration of the first graphical interface, hyperlinks, video conferencing, and mouse.

At this point, Microsoft’s approach to strategy presentations was adept at mixing a combination of BillG-style architecture slides with short and slickly produced video vignettes complete with keyed in screen mockups. The series of scenarios envisioning the future of software enabled by .NET formed the heart of the strategy articulated at Forum 2000. They featured the gamut of nascent technologies that were frequently talked about including tablet/pen computers, handwriting, wireless, voice control, video chat, location awareness, presence awareness, notifications, mobile devices, and so much more. The scenarios and designs had a decidedly consumer feel including the bubbly buttons and logotype used in MSN.

There was a slight problem in that the gap between what the audience saw in those demos and what any team might have been working on was, well, significant. It was not that many technologies were decades away from possibility, but only bits and pieces of a product were being worked. The role of a platform strategy is to inspire, however, not necessarily detail everything that is available in short order. Like the 1994 Information at Your Fingertips keynote, these sketches of the future were prescient and designed to create a north star for the company (a favorite expression) and even the industry. So while it was a challenge to be so far out, it was intentional. Unlike previous visionary presentations, Forum 2000 was far more specific in terms of product and roadmap. That was the problem.

Today many look at these presentations and the underlying products that emerged as evidence of many ideas where Microsoft was early, but for one reason or another fumbled in the transition to the modern world. It would be fair to say Microsoft was early to many shifts over the years, but as was so often the case being early ends up being wrong as well. When one is early and fails, the problem was usually that the culture or technology underpinnings are not yet mature enough to support the vision or the market was simply not ready. One could go through each of the technologies shown to realize the decade that would be required to bring them to market. Tablet PCs required screens and processors that did not exist. Handwriting recognition had been stuck at a level of reliability that was more frustrating than useful. Mobile devices would undergo a huge transformation with touchscreens and ubiquitous data connectivity. The services talked about would ultimately arrive with an entirely different architecture than Microsoft was building out at the time. Technologies such as XML would be widely used, but as commoditized as plain text files have always been and confer no real proprietary advantage as Microsoft hoped. Other technologies, such as virtualization that were key to the early cloud era had been rejected and would not play a part in Microsoft server strategy for another 5 years. Early efforts tend to be pointed in the right general direction, but the small errors or incorrect initial assumptions compound over time and ultimately diverge far too much from what eventually makes it to market.

Strategically, the comparison to the transition to the graphical interface was front and center and was used several times throughout the presentation. Our shared desire to repeat that transition and the success that followed provides evidence that being early is good. Windows arrived before the computational power and memory capacity could run the software we built. Taking time for technology to catch up did not deter us. Perhaps what would ultimately trip us up, however, was the grandeur and interconnectedness of our collective plans that left little room for execution error or for influences from the outside world and what was transpiring on the internet at a rapid pace. As many would note critically following the event, it was still about Windows at the center and to truly be a new strategy the conventional wisdom held that Windows needed to be abandoned. It is not at all clear to me that was the mistake, though it does make a simple narrative.

This was the peak moment for the catch phrase developed in response to the antitrust complaint of integrating software into Windows—integrated innovation. We overachieved on integrated innovation in that everything was integrated with Windows as we thought we should be permitted to do. Our defense was also our strategy, and also our technology foundation.

Nevertheless, the industry was excited by all this big talk. If there was a theme to the day, it was innovation. Every section of the day featured an explicit slide calling out innovation. This was a subtle jab at the critics and regulators who felt Microsoft achieved a dominant position and grown subsequently complacent. Innovation highlights were provided for .NET Services, .NET User Experience, .NET Programming, Small Business, and Business Users. That’s a lot of innovation! In addition to the videos, we showed live demonstrations of code: the earliest TabletPC prototype and handwriting recognition, a new browser-based service for small businesses, and the technology demonstration described earlier.

Perhaps more abstractly, the day was about a new era for Microsoft. There were the existing products and from this point forward there were new products built in new ways that would solve the problems the old products had built up. Everything was going to be faster, take less memory, reduce administrative burdens, and provide new levels of capability and convenience for customers. Microsoft clearly divided the world into old and new. That was bold and companies almost always lack the fortitude to make such statements clearly.

Competitively and concretely, .NET (using the term broadly as everyone did) was Microsoft’s answer to Java for server programmers. That was the big battle driving the platforms strategy. Java had captured the hearts and minds of developers building web server applications. The .NET technologies for enterprise software development would go on to create the platform that dominated in-house enterprise IT software, creating a generation of .NET programmers who today are more than comfortable with Microsoft as a provider of cloud infrastructure, even if it is Linux and not Windows. While the .NET programming tools would launch over the next year, this was the first real stake in the ground. On the PC desktop and client, we had won with Internet Explorer, which allowed the vision for the user experience portions of the presentations to move forward with fewer constraints and a focus on what was done on the PC but also supported in the browser—a desktop-first strategy.

It took 18 months before the first product release with the .NET architecture, Visual Studio .NET which was the first product to use the .NET name. The server product line underwent a pivot to support the new capabilities. Ultimately, .NET and its companion and proprietary programming language C# were enormously successful for Servers and Tools and came to define the era of enterprise client/server computing—so much so that most of today’s leaders in IT were products of the .NET era and, as a result, Microsoft created a generation of business IT leaders strongly connected to the company. Much of Microsoft’s strength today in enterprise accounts can be directly tied to IT leaders that rose up the ranks by betting on .NET.

Closer to home, the session on “User Experience” which was really about Office featured a presentation by a group building what instantly became known as Office.NET even though there was a clear demarcation of “technology demonstration”—we loved to think these small changes in wording brought us air cover or permitted distinction between products and directional demonstrations. To clarify, the technology demonstration did not claim to be Office.NET but the roadmap slide of product releases we presented at the time used the name and provided a “2002+” ship date.

The technology came from BrianMac, creator of Outlook, who upon leaving Outlook started up a new team called NetDocs, for network documents. NetDocs reported to BobMu, my manager at the time as well, though Brian and I had not crossed paths all that much since Outlook. We were both focused on what we needed to get done separately.

BrianMac formed the NetDocs team much the same way he built the Outlook team, growing the team to over one hundred in short order. The vision for the product was expansive and included many hot, new technologies. It was also being written in the latest technologies, including the latest magic technology XML (eXtensible Markup Language, which was becoming increasingly popular as part of programming for the browser) and, more importantly, it was using many of the new capabilities in Internet Explorer. XML was also the latest magic beans technology that took on capabilities much greater than reality. Brian had a knack for constructing expansive visions assembled with the strategic technologies as we saw with the creation of Outlook. As with Outlook these technologies were new, unproven, and unfinished. Outlook did quite well.

The scenarios enabled by the NetDocs vision subsumed Office, particularly Outlook, Word, Excel, and more, but with a decidedly modern take. By modern, the implication was that people no longer needed to worry about which Office app to use as there was one single document type, the universal canvas, that worked equally well with words, numbers, graphics, and email, and was easier to use because of that. This was not a new vision and in fact the idea of integrated packages had a history of attempts from both Lotus and Ashton Tate in the pre-Windows era, as well as Microsoft’s Works app (a modest success for price sensitive customers). The all-in-one application was a favorite among the first generation of PC users and BillG in particular who routinely complained about overlap and redundancy across the various “modules” in Office, modules being his favorite way to describe an app in the suite. Would this time be different? Did the processing power and memory finally enable this? NetDocs set out to prove it could.

I was skeptical, but from my position in Office skepticism was viewed by others as defensive and territorial. I wasn’t being defensive. I just didn’t think it could work. Others projected it could be a $1 billion business within three years.

A running joke at the time was that every new product idea somehow included digital photos, and Forum 2000 was no exception. Every demo included digital photos in some form. Digital cameras were the hot consumer item and sharing photos on the internet was becoming mainstream. NetDocs not only included photos, but to illustrate the importance of photos, a product that was extremely innovative but languishing at retail without sales and marketing support, Microsoft PhotoDraw, was reorganized into the NetDocs team. This method of building a new team by acquiring other internal teams and jettisoning their existing product was a strategy employed with Outlook as well. I was a big fan of PhotoDraw and to me it was one of many examples of innovative tools Microsoft created that we were unable to capitalize on because it was too small on its own and too niche to be part of Office—this will become a familiar theme shortly.

There was another running joke that every new product idea being dreamed up somehow also included electronic mail—email was the anchor of the internet and became a big deal for AOL, Yahoo, and MSN. Microsoft was a clear email leader with Exchange, but that was for business. For consumers, Microsoft’s MSN division acquired Hotmail in 1997, the first web-based, viral, and free internet mail. The number of email users on that service was approaching 100 million, when the entire internet population was roughly 300 million. NetDocs also became email. That should not be a surprise given the roots of the team and leaders.

Photos, email, calendaring, XML, word processing, spreadsheets. . .that’s a lot, a lot to like.

NetDocs also enabled a new subscription business model. There was nothing particularly technical about doing this work, though convincing customers to rent software (as people thought of it at the time) was new. The team was working on a new technology to provide seamless updating of the NetDocs Windows desktop application over the internet. Seamless updates might convince customers of the benefits of rental over ownership as the product could be enhanced without purchasing anything new. Customers were struggling to deal with updates, most of whom were not yet able to use the Windows Update service that is now standard on every PC.

Given my early experience getting the first version of Outlook to customers, I remained skeptical of NetDocs achieving all that was sketched out, especially without the kind of constraints being part of the Office release imposed on Outlook originally. The amount of code to write, the ever-changing scope (and resistance to constraints), the huge challenges of building some compatibility and interoperability with Office, as well as the fragility of the technology foundation—as the latest and greatest always seem to be—were not usually a recipe for success and seemed familiar. The success Outlook achieved, due in no small part to being free and bundled with Office and more importantly the only client for Exchange email, provided a halo of sorts for NetDocs. This is a good lesson in how success in a big company can take many forms beyond customers laying out their cash for a product.

I had not paid attention to NetDocs (nor NetDocs to Office) and now suddenly and without warning, NetDocs was front and center strategically for the company. NetDocs was filling a void in the strategy, at least internally which was that for the .NET platform to be successful it needed a killer Office application. I should have internalized that strategic point going to the NGWS meetings, but I did not. I managed the Office10 project aware of the costs of choosing to lay low—that people would view Office as failing to support the platform or even to acknowledge the future. In an industry (and especially a company) where the next version is always way better than the current version and new platforms always require the leading apps to support them, it was challenging to take this approach.

This old versus new dynamic always creates tensions in a large company. Echoing Innovator’s Dilemma (again), a series of press stories played out over several months. While there were grumblings, overwhelmingly people on the respective product teams were not consumed with potential overlap—Microsoft had a long history of next generation projects that fizzled. When will NetDocs replace Office? Will Office stand by and allow NetDocs to replace it? Will customers be confused? How will the market deal with two kinds of Office products? This was a far cry from the Cairo versus Windows NT, or the Windows NT versus Windows 95 battles that played out over years, at least I thought that to be the case.

During these times the negatives the market perceives of the incumbent are amplified irrationally—software bloat, nothing left to add, slowing growth in the business, and more. Simultaneously, the perceived positives of the new product are amplified irrationally: sleek, modern, simpler, faster, lighter weight, innovative, and new. Microsoft was great at setting up this dynamic. I had been the poster child for old technology and resistance to change more than once (Java Office, component Office, web Office), and while I could brush it off, in the case of NetDocs and Office there was quite a bit of bashing externally of a product that was half of the company’s profits. The internal tension was significant, but not because of a deliberate product competition or organizational competition for resources, but because there’s no way to constructively align the past Office with an ever-expanding vision. Regardless of the strategy, NetDocs could have laid low and first spent a couple of years building a product. It just wasn’t in Microsoft’s culture to do so at this time given the demands to put a big vision out there.

Nothing could stop the Forum 2000 train. This was exactly what NGWS needed. There was broad satisfaction with the event, even though ongoing legal challenges clouded the strategic presentations and strategy. The future was .NET everything, including Office.NET.

Deciding to show NetDocs at Forum 2000 was controversial, at least with me, and probably not many others. I was usually the most conservative about showing products or features with an uncertain path to shipping, let alone version 1.0 products built on version 1.0 technologies accomplishing new scenarios that often didn’t pan out. There were also legitimate concerns that word of a modern Office.NET could slow or halt progress on enterprise agreements, in an extremely touchy post-dotcom bubble business environment.

After many email threads prior to the event, NetDocs ended up showing some basic features, such as typing into a word processor-like screen, summing a column of numbers without launching a spreadsheet, and a calendar scenario that used XML technology to merge a personal calendar with a Seattle Mariners calendar. The session painfully reiterated the “technology demonstration” aspects of the demo and never used the phrase Office.NET, though the prominence of Office.NET on the roadmap left few dots to connect.

To mitigate the risk to enterprise agreements, the demo was said to be relevant to Microsoft’s new small business offering, briefly called bCentral. For almost another decade, the Office strategy for the web and internet targeted small business starting with bCentral and always using branding to show a distinct separation from Office for enterprise. This compartmentalized a new approach to the less risky market segment where Microsoft had more upside than downside. For big business, they were pushed to see things through the lens of Windows Server and the software housed in company data centers along with desktop Office, all available with an Enterprise Agreement. This was a defensive approach but was consistent with how customers thought about Microsoft products. Word and Excel were indispensable tools for small business, and increasingly Outlook, especially with many add-ins, was the preferred tool for small businesses to manage sales and customers. Customers could not buy Exchange and set up Windows Active Directory and file/print servers fast enough. In practice, operationalizing the transition outlined was more practical and somewhat defensive, echoing the Innovator’s Dilemma.

The Wall Street Journal was quick to pick up on the potential challenges as well in a story by Rebecca Buckman shortly after the event “Microsoft Readies a New `Office' While Renovating the Old Standby” where she wrote:

What does a company do when its single-biggest product is in danger of being eclipsed by new technologies?

If that company is Microsoft Corp., and the product is Office, it sets up a stealth team of crack engineers to dream up a brand-new version of the software suite -- while continuing to crank out the old standby. It's a tough, two-track strategy that has pitted a "today" development team against a "tomorrow" team, as one person close to the teams puts it -- and it's still unclear, he says, "how today and tomorrow will meet."

As we know now, the danger of being eclipsed did not materialize. We deliberately and somewhat nervously made that bet, as previously described. The teams were not pitted against each other, not yet anyway.

PCWeek reporter Mary Jo Foley loved the NetDocs story and wrote about it many times. By the end of the year, a few months after Forum 2000 she said in a story “Netdocs: Microsoft's .Net poster child?” where she described the product as:

Netdocs is a single, integrated application that will include a full suite of functions, including e-mail, personal information management, document-authoring tools, digital-media management, and instant messaging. Microsoft (msft) will make Netdocs available only as a hosted service over the Internet, not as a shrink-wrapped application or software that's preloaded on the PC.

Netdocs will feature a new user interface that looks nothing like Internet Explorer or Windows Explorer. Instead, Netdocs will deliver an integrated workspace based on the Extensible Markup Language (XML), where all of its application modules are available simultaneously. This interface is based on .Net technology that Microsoft, in the past, has referred as "Universal Canvas."

There was nothing sinister or even that playful about what was going on. It was just. . .going on.

We were busy, and well into building Office10. At the very least, until NetDocs was usable (self-hostable) by more people, the best thing to do was let them keep writing code and hope they stopped trying to recruit people from Outlook.

NetDocs versus Office was just going to simmer for a while. There was no way around that. Because we were in the midst shipping (eight months to go) and NetDocs now had a deadline of sorts, there wasn’t room to attempt to reconcile the products, have them relate strategically, or really do much of anything. I was happy to work heads down and not worry about it.

Putting Forum 2000 in perspective, SteveB sent an all-company follow-up email briefly laying out the strategy and timeline we were undertaking. He reinforced the huge change by referring to the strategy as “Microsoft .NET” (the space after Microsoft was important), coming very close to rebranding or even renaming the company around this new strategy:

Microsoft .NET will be delivered in three forms: a new user experience; infrastructure and tools; and a set of programmable .NET Building Blocks. This is a long-term strategy, one that will take years to execute fully, so it is critical that we all stay focused - not only on our goal but also on the daily steps it will take to achieve it.

The Office team was focused on Office10. We would worry about NetDocs later.

On to 065. SharePoint: Office Builds Our Own Server

PS: Many readers lived through Forum 2000. Some have shared their own experiences from the event, like this wonderful post from Charles Fitzgerald, Exploring Alternative History. Please share your experiences on twitter or in the comments, especially if your personal experiences bring a different perspective as they well might.

Some additional moments from the Forum 2000 video:

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit hardcoresoftware.learningbyshipping.com