055. Office 2000 is Good to Go!

Office9 (aka Office 2000) was the very first release of Microsoft Office built by a team that began the project and ended the project as one. Well, almost as there was that pesky Outlook challenge. That said, the project was coming to a close by the end of 1998 and it was late. It was not terribly late or out of control as I reflected on my 1:1 with SteveB, but late nonetheless. The way old-timers would talk about projects to me was that product development slogs “really sucked” and then you ship and that is the “best feeling ever” and it is worth it. I was hoping to feel that way soon. The last month of 1998 would be eventful for me personally and the team in ways that were unexpected, both happy and tragic, and a reminder of the scale at which we worked.

Back to 054. Steve and Steven Get New Jobs

Not only were we five months late by the end of 1998, we also faced the lag time that would happen on the other side of our pending a holiday break. It became clear our end-of February completion date for Office9 was going to be a challenge.

At least we were on the glide path to completing the product before we left for the holidays. As far as the enterprise sales team was concerned, we were going to finish in the first quarter, which was important for customers buying or renewing volume license contracts. Office 2000 was shaping up to be a significant release for the end of the millennium. And yes, Office 2000 was Y2K ready.

As I prepared to head to Florida for holiday vacation to see family, JonDe let me know that I was being promoted to corporate vice president. This was truly a big deal personally, and as I look back I realize it was a big deal for the company. Along with BrianV leading Exchange a press release would go out announcing our promotions in about a week—the ranks of executives were small enough then that not only did we do press releases for promotions, but we became named officers in the company and the two latest additions from the product development teams. If you’ve ever been part of a rocket ship, what you experience is growth in executive promotions when you start to wonder if this is a worthy promotion or will people outside your immediate group roll their eyes. NatalieY, the spiritual and cultural leader of the company in human resources, sent me an incredible note telling me how “real” the promotion was which meant everything. These insecurities come with what does seem like one of the most significant milestones in a career and for me, I will always think of this as the most significant, owing much to JonDe and all of the Apps people before him.

While everything was the same organizationally, what was simply a title change in the Exchange address book was the start of being treated differently, particularly by those I did not yet know, especially in the field sales org, and, perhaps even more entertaining, by the various Microsoft systems and services.

In short order, Colleen Johnson (CollJ) was being briefed on the exec travel desk (EXTRAVEL), executive tech support (EXECSUPP, yes, there was special on-call PC and tech support for executives), executive shuttle (sort of an Uber for VPs to go between buildings that I almost always walked anyway), and even the special team that helps make PowerPoint slides for execs, and so on. I knew these existed, but I thought they were for Bill, Steve, the CFO, and Board, not all the VPs. Oh, and my card key worked in all sorts of places it used to not work, like weekend access to the Executive Briefing Center! I did have mixed feelings about special treatment for run of the mill execs in product groups. The teams were all incredibly nice, but this felt like excess to me.

CollJ was a key addition to our team who had an outsized impact on our operational excellence and culture. Hired as an administrative assistant, she quickly and quietly assumed the role of indirectly managing the dozens of administrative/group assistants across the team (on average one for every 100 people) which was the front line of our culture of careful management of resources. She also took on the role of formalizing all our headcount tracking which was a key enabler for how I would manage the team. A dumb example was instituting a code in our SAP-based system for what individuals worked on (Word, Excel, Office shared, etc.) and what functional group they were in (dev, test, program management). Amazingly our systems did not track that and yet it is all that mattered in managing a significant team. For decades neither she nor I managed to coach other teams into managing this way, even after the yearly visits from teams being told to learn a best practice from how we managed things. There’s a lesson in there—you can’t manage a big team at scale if you don’t know what people are working on and how humans (not dollars) are allocated!

Colleen also led our efforts at developing a new more scalable Office culture. Too much of Microsoft was still acting like it was college when it came to team outings or now operating like a fancy Wall Street bank when it came to using company resources for events. From figuring out how to get a huge discount on team movies by renting all the theaters for morning shows and offering family friendly and alternatives to the current sci-fi release, to helping teams to know about $250/day offsite locations (including a stop at Fred Meyer for snacks) instead of the $2,500 locations (before the required hotel catering), Colleen instilled a broad sense of fiscal and cultural responsibility across the team of admins. She helped to create whole classes of events such as our product vision meetings, the “tradeshow”, and especially our launch events. As the team grew, the tradeshow became a hallmark event where every member of the team had a chance to experience all the other teams building Office as though we were at a tradeshow. Years later, Microsoft Research would adopt this format for what became the Microsoft TechFest event.

As a VP I quickly learned people I did not know would no longer email me directly and even people I did know would soon also employ the level of indirection afforded by CollJ. The Microsoft culture had developed an arm’s length VP culture, where contacting a VP meant going to the Exchange address book and looking up the direct reports of a VP to find their Executive Assistant to contact about “getting time with” or “what is the best way to email them”. Exchange had a unique feature where you could offer an assistant direct access to an email account as a delegate. Suddenly VPs were having admins screen email and even respond on their behalf. How quickly we became “big”. Along with helping people to email me directly, Colleen had to remind people of one other quirk of mine I insisted on, which was I managed my own calendar. I was hardcore about this because of what I’d gone through in terms of the ripple effect of scheduling. A meeting was scheduled with an exec, the executive assistant moves the meeting for something important, then everything dependent on that moves as well. Soon the one person who needs to get out of the way, the VP, is the barrier to making progress as a routine course of business.

I think I spent the rest of my years trying to hold on to a feeling of small, and avoiding the distant feeling new hires (and we had hundreds every year) would have towards execs. I think I had varying degrees of success at doing so and certainly made my mistakes at trying, but with Colleen’s help I worked hard at that for my run, even as we scaled.

The press release went out while I was on the way to Miami. I was expecting an uneventful time in condo haven, North Miami, aka Del Boca Vista. Working to avoid the tourists in the city, I was reading the pre-holiday Miami Herald newspaper (scoping out a holiday movie to see). The paper contained a story of tragedy at Disneyland. Anything to do with Disney received a great deal of attention in Florida and caught my eye having grown up in Orlando. A guest was seriously injured on a ride, through no fault of his own, and later reporting said he died on Christmas Eve. The first coverage did not have a victim’s name but soon the details emerged.

Almost at the same time as the first story, an email arrived from Jeanne Sheldon (JeanneS), who was leading Word testing at the time, asking me to call as soon as I could, which was not routine. Over the phone, I learned from Jeanne the victim was a senior test engineer on the Word team who, like me, also started at Microsoft in July 1989, after first emigrating from Vietnam to attend university in Paris. His wife was also injured though expected to recover. Their son escaped injury. JeanneS took it upon herself to craft a note that was to be the first email I would share as a vice president to the entire division. The subject line read, “Sad News.”

Microsoft was still young enough that we did not have in place the big company processes that eased these tragic situations. Much of the grieving happened in email over the holiday. It was an awful time for such an awful tragedy. I was starting to learn that at a certain scale, every kind of life experience, joyous and otherwise, would be part of our team. Microsoft had a few sad times before, and even some close to home in Apps, but this was difficult in its own way.

As we returned to work the project was winding down and we were in bug fix mode where only critical bug fixes were “taken” meaning only the most serious bugs were addressed with code changes.

Despite being a team always worried about engineering productivity, we were in that phase of a major software project where 2,000 people came to work every day and basically did nothing, if the measure of something was making changes to the product. Testers ran and re-ran tests. Developers investigated bugs and decided if the code change risk was greater than the risk of leaving a “bug” in there. Program management was fielding endless inbound requests for just one more thing or digging into one last potential oversight. Documentation and Localization were working on producing international releases. Lawyers were combing through marketing materials and documentation, and of course adding more words to the end-user license agreement (the EULA). We were all using the product as end-users and testers on every computer we owned.

The biggest disappointment we were having about the new release was the lack of excitement within Microsoft. Whereas people were beating a path to the servers to install Office 97, we struggled to get Office 2000 deployed in large numbers across the company. This was, in reality, a sign of changing times.

Broadly pushing, perhaps forcing, internal use prior to shipping was another cultural difference between Apps and Systems. The Systems view was always hardcore—hardcore about pushing internal use, sometimes even too early, and hardcore about not using competitive solutions. The first was enabled by the long end game of shipping a Systems product, for example, Exchange Platinum (Exchange 2000), which began use inside Microsoft in 1998 and did not ship for almost two years. SteveB even apologized at an all-Company meeting one time for the bumpy Exchange pre-release. It was in beta for most of the entire Office 2000 product cycle. Competitively, Systems often rooted out competitive products and made it a goal to remove them (like Oracle server or later Google search).

The Apps view was always a bit less aggressive. The time from the product working until shipping was much shorter—there was less time when the product was usable by the typical employee, and by then a typical Microsoft employee was not much different than a typical employee in most any large company. We always viewed the use of a competitive product as a failure on our part, but one to learn from not to force away. If an individual or team wanted to use an alternative, then we would not object but want to understand why. The biggest example at the time was the growing use of Adobe’s PDF instead of using the native file formats that BillG had insisted upon.

Our testing and release process did not rely on an extended period of internal testing or external beta tests. The nature of our products and process enabled us to achieve a high level of quality, even during these maturing days of the PC. Perhaps this was misplaced confidence as there was little data to base this on, but we closely tracked support calls and enterprise customers to have a good “feeling”. In the next chapter we will have an eye-opening experience when it comes to data informing these decisions.

In the case of Office 2000, we were starting to see a sea change in Microsoft and the industry. Office was not the only place (or even a place) for excitement at the time. Browsers were really exciting. Consumers were excited by new MP3 players, not laptops. Most of all, enterprise IT was not excited by anything that caused them work—their cycles were being used trying to stabilize internal infrastructure, convert legacy client/server to the web, and prepare for Y2K.

We were losing competitively but to different competitors, not the alternatives to Office we feared.

The biggest competition for Office 2000 was . . . Office 97.

We were so heads down finishing Office 2000 that we didn’t realize how well received, and how good, Office 97 was. Everything we announced at our Office 2000 enterprise event in New York was solid and the feedback from the beta was good, but we faced resistance to upgrade because it took work. While customers already owned and paid for Office 2000 with their multiyear agreements, the cost to deploy (the cost of change, support, labor, etc.) had to be considered. To deploy Office 2000 or not was a major decision point within IT.

The sign-off for Office 2000 took place on a sunny April day in 1999. The product was eight months late from our original schedule we picked in March 1997. Unlike Office 97, however, the team was not frazzled, but tired. We faced the complexities of pulling everything together, but we improved the process and came together as a team. Still, an eight-month error is big on a 24-month schedule. We would conduct a detailed postmortem and make a series of changes.

CollJ and the admin team commandeered the fountain area between buildings 16-17-18. The makeshift stage and a megaphone were ready. Continuing with the theme of a maturing culture, Colleen instituted limits on alcohol and everything went smoothly except for a minor champagne incident inside Building 17 that the Art Committee was rather upset about. We grew up a little bit more this ship day.

Earlier in the day we met in the ship room, with one representative from each of the 20 or so teams. Going around the horn like Mission Control at Cape Kennedy (1999 was the 30th anniversary of the moon landing so space was everywhere), we proclaimed Office 2000 “ready for the web” and “good to go.”

I missed signing off on Office 97, but this time, as the VP of Office, I was going to get the “special treatment,” meaning I was going to get thrown in the fountain. To everyone’s surprise, I came prepared, wearing a bright yellow rain suit and goggles (to protect from flying corks). CollJ printed out a giant copy of the paperwork that went to the manufacturing plant that duplicated DVDs and we did a ceremonial sign-off with BillG, who made a fast getaway to avoid the celebration.

The surplus air raid siren, a desktop applications cultural touchstone, sounded (though technically illegal to set off in the city of Redmond), as was DAD tradition for every ship party and sign-off. The next thing I remember was sitting in a fountain soaking wet.

The events of the day were memorialized with a videotape (an actual cassette), which each member of the team later received, including a congratulations from BillG at the end after rolling the credits for the release. Almost 2,000 names scrolled by while the Office Assistant PowerPup looked on. Still always looking to save money, we didn’t do anything fancy for the credits—it was a Word 2000 document that I scrolled using support for the wheel mouse (introduced in Office 97).

Our growing business with enterprise customers and the arrival of the internet introduced a new step in releasing the product. Enterprise customers would now begin deploying Office 2000 right away and would not have to wait for the retail arrival of boxes. We announced RTM for enterprise customers with a press release. In a few weeks we would actually launch the product for the retail market with a global series of events. I was off to Japan.

It wasn’t just mission accomplished. It was my first mission as a general manager and executive, and it did feel different. Standing on the makeshift stage in my protective gear and signing off on the product—the first release of Office as a single team—was an emotional product moment.

On to 056. Going Global . . . Mother Tree

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