At one point, Allard brought Jon Thomason in to lead the Xbox Live team. His operating system group, which he still nominally led, was still cranking out monthly releases, but Thomason’s involvement was considerably reduced. “There wasn’t much architecture work to do, management was pretty simple. So I went to work on Xbox Live.”
During the first week on the project, Thomason went around talking to all the principle people to get a good picture of what was happening in the project. “I was a little in the dark about what they’d been doing.” When Thomason realized that not only did they not have a launch date, but they didn’t even have a clear schedule, he announced that they would have a complete schedule in two weeks. “So I set a date for a date, which I know is kind of ridiculous, but it had to be done. And we went and decided when we could be done and set a launch date.”
Thomason took his schedule and launch plan to Robbie Bach and the other VPs. “I said, ‘We can be done this day, but we want to set a launch date right now and get all the clocks ticking towards that. And I got them to buy off that day.” Launch day was November 15, 2001. That day, we set that date, and we didn’t change it after that. We hit that exact date.”
Marketing: Four Points of XBL
Thomason’s team then went into high gear, adding some new people and working on the data centers and other important components of the system. The team had grown by this time to 60 or 70 people, but from Thomason’s point of view, the weakest link was the marketing team. “The marketing team that was working on this was extremely junior. So not only did I do all the engineering stuff, but I ended up doing all the marketing planning for Xbox Live, as well. I actually wrote the marketing positioning for Xbox Live.”
Here are the four marketing points with Thomason’s comments:
One: Friends List: “Across all games, see if people are online no matter what they’re doing on their Xbox”
Two: Voice Chat: “Talk with any gamer anytime, no matter what they’re playing.”
Three: Gamertag: “Your online identity and reputation across all games.”
Four: Online Multiplayer: Quick match and optimatch. “Play with your friends or play with people your skill level.”
“We wanted to make sure that we were clear about what Xbox Live was all about in terms of features, and how better to do that than to emphasize features that we had that Sony did not.”
“I remember one meeting where I had all the VPs in there—Ed Fries was in there, and of course Robbie and J and everybody—and I remember, I made these guys memorize the positioning. I had four positioning points that we were going to talk about, and I made them actually recite them back to me because I didn’t trust them (heh), and I remember Ed sitting there, ‘OK now. Let me say this again.’ He said it three or four times, and they all got it, and it worked. Because when we launched that thing, if you looked, you found all those four points in every single article. And I think we nailed it.”
Two Styles
Although Thomason credits Ferroni’s leadership for getting some things done, he personally thinks that the whole project could have been completed within 18 months, and it had been 18 months by the time he took over. He believes that they could have been doing more work in parallel with the system development and that they didn’t hire the necessary people fast enough.
Multerer sees it somewhat differently, however. He agrees that it might technically have been possible to launch XBL with Xbox. He says, “Jon might be right… we could have shipped something by the time Xbox shipped, but it wouldn’t have been as compelling as what we did ship.” He goes on to say that what they did ship—System Link—“was a precursor. Now if you want to build out a social network, especially when there are no social networks to understand from, I’m sorry that takes more than nine months to do. System Link just barely got done.”
Multerer worked alternately under both Ferroni and Thomason. In fact, every several months the situation would change. “I was working for Cam. And then, like nine months later I was working for Jon again, and then a couple of months after that I was working for Cam again. And it just went back and forth and back and forth. I did not have a consistent manager for more than a year during that entire formation.” He saw both of them as effective, but in different ways. “Jon is really good at shipping. Cam is really good at figuring out what to ship. They have very different approaches. A ton of work got done while Cam was running it, and yeah, there were lots of distractions. So I’m not saying it was super-efficient.”
Thomason spoke about his philosophy of development, which was on display when he ramrodded Xbox OS development. “You’ve got to be ruthless. You’ve got to cut stuff. The big decisions were actually setting a time, and then cutting things to make it work. There were millions of little decisions, but the big one was setting an end date and turning it on a schedule-driven project. That’s what they didn’t do, and that was the missing piece. They hadn’t said, ‘We’re going to ship this thing at launch, come hell or high water,’ which with the system software, we had to. We had no choice. So I treated the Live launch like the same thing. Within the first two weeks we had that launch date set and bought off on by all of upper management… everybody.”
Thomason says that making an official deadline made everybody work a lot harder, but adds, “I didn’t do it by flogging everybody and doing the Office Space thing, you know… saying he’s got to come in to work on Saturday. It was more like, we’ve got to make this date, and to do it we had a work back schedule that got us there, and then people worked weekends because that was the only way to do it. We had to get everybody into a full-on, heads down, mission critical sort of attitude, and they hadn’t done that previously.” Thomason says that they didn’t hire early enough, but by the time he got there the team was pretty big. “Remember, the system software team was only about 20 at that time.”
Thomason does give credit to the work the team did, especially Boyd Multerer and Dinarte Morais. “Dinarte Morais wrote all the security software for Xbox Live. He wrote the stuff that copied code down to the console and then ran chief detection code on the console, which at the time was unheard of. Nobody had ever done that before. Everything else had been done on the server before, so you could hack it by hacking the client, but we had a secure environment… secure enough that we could actually push the code down, while it was authenticating the login, and if it didn’t look right, he could just fail it. It’s pretty cool, and it’s worked. There have been a lot of Xboxes over the years, right?”
Some Things Take Time
Looking at the project the way Thomason did, it seemed clear that the project could have been completed more quickly and with better planning. But Multerer points out the many challenging technical aspects of XBL, including making solid peer-to-peer connections; developing the security and exploit measures; designing the directory services, including matchmaking, leader-boards, a lot of value-added services, messaging systems, invitations systems; and user interface design.
“In fact,” he says, “the high level vision of the project didn’t fully solidify until the end of 2000. The vision starts to move into strategy. There’s the pyramid: Vision is at the top, then comes strategy, and then comes tactics, and I’d say Gamertags were at the strategy level.”
And that was another thing to keep in mind. It took time to promote the more controversial aspects of the system, notably Gamertags and their concept of presence. They had multiple people and groups doing outreach to developers to familiarize them with these new ideas and try to get them onboard. He specifically mentions several names, including Ferroni, Scott Henson, and George Peckham. “And Robbie was all behind it,” he adds. “He stuck by us in that one during an extremely controversial time.”
He admits that there was a certain amount of tension between Ferroni and Thomason. They approached product management very differently—in a sense like the difference between a creative style and a more military approach.
To Multerer, both Ferroni and Thomason contributed necessary skills to the project. “Jon is extremely good at shipping things. Very, very good at get-it-done. Spends less time on ‘are we building the right thing?’ But even today, some of us who’ve been around a long time will refer back to Jon and, ‘Yeah, he was a master at shipping.’ But less so on the why are we building what we’re building.” The “why” came from Ferroni and Allard.