tylerbu @msft

12-year Doughnuts

As promised earlier today, some anecdotes from the past 12 years…

The Beginning

I interviewed at Microsoft in February of 2005, and started the following month on March 28. NEO was two days at that time, so I didn’t actually see anyone from my new team until the 30th. I worked on what was then called the Content Management Server team, which was largely composed of folks who’d come to Microsoft as part of an acquisition of a Vancouver-based company called NCompass Labs. Many of the engineers and PMs I worked with then are still working in ODSP or Yammer in some capacity.

My first manager was Jim Masson, who’s now the GPM of Word. Jim was a great managerial fit for me, and I learned a ton of what little I know about being an effective PM from him. I started right after M2 of the Office 12 (I think?) release cycle, and so I had a lot to ramp up on when I started. One of the first features I took on was called “Content Deployment.” That feature will come back in a later anecdote, don’t worry. 😊

Anyway, at some point very early on Jim scheduled a meeting with the devs, testers, and dev/test leads for Content Deployment, and they walked me through the feature. I didn’t understand any of it. I listened while they drew a bunch of diagrams on the whiteboard, explained everything and answered my dumb questions as patiently as possible, but it was a tough slog through some pretty deep technical stuff that I just didn’t understand. But nonetheless, this meeting is one of my best memories of working with Jim, because he sat with me in the conference room for over two hours after the meeting was over, late into the evening, and helped me fully transcribe everything from the whiteboard and attempted to make sense of it all. One of the things I learned from Jim is that in order to be a true leader, you need to be ready and willing to do whatever you’re asking others to do yourself. Jim sat through that meeting and beyond, long after he needed to. He stayed late because I stayed late. That meeting has really stuck with me.

Jim also taught me the power of encouragement. There were definitely times that I questioned my capabilities as a PM, and I struggled with self-doubt for much of my early career. Jim was instrumental in helping me build my confidence and ‘find my voice’ as a PM.

Strug Track

Since I was a college hire with no friends or family in the area, I quickly became friends with another college hire PM on the team, Benoit Schmitlin. Ben didn’t have a car at the time, so we became friends initially because he needed to get some furniture from IKEA, and I drove him to pick it up. I won’t go into details on how we secured the furniture to my 1986 Nissan Maxima. I don’t think the statute of limitations has expired on that particular misdemeanor. 😊

Ben’s no longer at Microsoft, but much of my early career was spent in his office. We went to lunch and dinner together almost every day, and would come back to the office in the evenings after dinner to work for a few more hours. About a year or so into my career, Lincoln Demaris (he’s still a PM over in ODSP) joined our team, and our duo became a trio.

I have a treasure trove of good memories from 2006-2008 or so, mostly because I was working with these guys, and we made it fun. All three of us seemed to alternate between feeling like we were hot shit and like we were the worst PMs ever, and none of us knew what we were doing. We heard rumors that Microsoft had an ’executive track,’ where some particularly talented new hires were ‘groomed’ for management or more executive roles. We all wondered why we weren’t on such a track (in reality we knew why we weren’t! 😊), so to compensate our egos, we created our own track – the ‘strug track.’ The ranks later unofficially expanded slightly to include Tim McConnell, Tomasz Tomko, and Chris Richard (the sole engineer in strug track, but in many ways the quintessential member).

The Jimtionary

I think in your early career, your direct manager sort of feels larger than life. As you work with someone for a long period of time, you tune in to their personality or behaviors. Jim was a major character in that chapter of my life and career, and over time I and the rest of strug track created a number of words to describe what it was like working for or with him. This became the Jimtionary – a dictionary of Jim-influenced words. All the words are portmanteaus. For example:

  • Jim’s office was the Jimnasium.
  • A 1:1 with Jim was called a Jimboree.
  • A meeting with Jim and other people was called a Jimdig.
  • When Jim asked you to do something impossible for anyone but Jim to do, it was a Jimpossible task.
  • When Jim needed you to do a rather menial task, you were a Jimpanzee.
  • Plus more, but the quality declines quickly. 😊

I don’t remember who among us coined all of them, but we had way too much fun with this. I don’t know if we ever told Jim about any of this, and hopefully he won’t be upset to learn of it when this email is inevitably forwarded to him. 😊

MikMort’s Legacy

The Content Management Server team was kind of a sister team to the Windows SharePoint Services team (WSS). Well, really, the WSS team was the “platform” team that owned core SharePoint, and my team was building on top of the SharePoint platform. As I recall, Mike Morton was a PM lead in WSS when I started, and one of the things he owned was the “Visual Blueprint.” With so many different teams contributing to the product, there needed to be someone driving consistency in the UX and visual appearance across the product. That was Mike. Well, Mike and his team, but from my vantage point, Mike himself was the Visual Design Czar. If you wanted to make changes to the way something looked in SharePoint, you went through Mike. And Mike, well… Mike was probably going to say no.

I distinctly remember thinking that all this stuff was very draconian. Why don’t they trust us? I thought. And so my impression of Mike was that he was this control freak, which still makes me laugh now that I work more closely with him. As I’ve progressed through my career, though, and seen what happens when you don’t have someone filling the role that Mike filled in WSS at the time, I’ve totally 180ed my feelings about this. In particular, in Office 14 I think SharePoint did away with the Visual Blueprint, or at least didn’t do it the way Mike did. And in my opinion, if you compare the UI/UX across the SharePoint product between 2007 and 2013, you can see the difference Mike and his team made in 2007.

I share this story partially because I still find it funny how I had this almost mythical image of who Mike was based solely on this sliver of interaction early in my career, and that image was not even close to accurate. There’s a lesson on first impressions there somewhere. But I also share it because it reminds me that a lot of what a good PM does isn’t obvious until you no longer have a good PM doing it. 😊 A PM’s value is often only appreciated through their absence, which is sort of a shame, but c’est la vie.

Starstruck

One of my favorite things about working at Microsoft is the opportunities I’ve had to meet some of my ‘heroes.’ At my very first PM meeting, I got to meet Ward Cunningham, who developed the first wiki. I got to shake hands and chat briefly with Dan Bricklin, the creator of VisiCalc, while representing Microsoft at a content management conference in Boston. I had email exchanges with Chris Wilson, then the platform architect of IE, about web standards and HTML and CSS, and why XHTML 2.0 wasn’t the future of the web – HTML5 was (for the record, Chris and I agreed on this point 😊). These were really cool experiences, and helped me feel like I wasn’t just a guy from a relatively no-name school who was trying his hand at the software industry. I was a part of the industry, meeting people whose names I’d learned about in school or through the news. Early in my career, these experiences made it obvious to me that I, too, could have the opportunity to make a dent in the universe through my role at Microsoft; I just had to make the most of those opportunities.

Dogfood

When I started the Office 12 M2 dogfood build was the new hotness. Hot it was not. Here’s a graph I drew in OneNote on August 24, 2005, almost 5 months after I started. As I recall, the M3 build was released around this time, and I was appalled that it had gotten even worse than M2. This graph is based on anecdotal data, but as we all know, the plural of ‘anecdote’ is ‘data.’ 😋

The Dogfooding experience was so bad for me, particularly in Outlook, that the Outlook team sent an intern to basically shadow me and see how I used Outlook, because I was classified as a ‘screwed user.’ Apparently there was some small percentage of users that accounted for something like 70% of the errors reported to Outlook, and the team was trying to figure out what about these ‘screwed users’ actually screwed them. So this intern followed me around with a notebook taking notes as I used Outlook. It was one of the weirdest experiences I’ve ever had. I never got the results of the report, but I did get to switch Exchange servers like a bazillion times. Fun!

Below is another poorly-drawn comic depicting my dogfooding experience. I made it while I waited for Outlook to boot back in 2005. While I took some hyperbolic artistic license, it would literally take hours for Outlook to boot. If you ever wonder why I can have such a bad attitude sometimes when it comes to dogfooding, this is why. 😊 Outlook 12 M3 dogfood ruined it for me.

Tailor Bulter

Speaking of bad experiences, back to Content Deployment. 😊 Content Deployment was a really rough feature. It was ambitious in its scope, and core to its functionality was a technology in SharePoint called PRIME. I don’t remember what all the letters stood for, but the I and the E were for Import/Export. I think the M was either Migration or Move. It doesn’t matter. The only important thing to know about PRIME is that it was (is?) incredibly complex. Note that I didn’t say it was complicated – it was complex. It was attempting to solve what on paper seemed to be a simple problem, but in reality was about as far from it as you can get – that problem was copying/moving/importing/exporting content and metadata from a SharePoint site or web to another SharePoint site or web. Would that it were so simple.

Content Deployment was one of the biggest users of the PRIME API, so over time I became the ‘PRIME guy.’ There was just one problem. PRIME was absolutely riddled with bugs, and even better, it exposed other people’s bugs. By exporting and then importing content using PRIME, weird things would often happen. Oh man, and if you imported content from an older build even weirder things could happen. Guess who got to evaluate and triage every single one of those bugs, even though many of them ended up not being PRIME bugs? That’s right. Then I got to argue with other folks about why the bug needed to be fixed by them because even though PRIME exposed the bug, it wasn’t a bug in PRIME. Good times.

Anyway, a few months after Office 12 shipped (lol, you thought people used our software soon after release? Nope.), the bug reports started pouring in. I tell people it was the #1 QFE/escalation generator in SharePoint at the time, and no one has ever tried to dispute that claim, so I’m sticking to it. Eventually things got so bad that I stopped working on Office 14 stuff and basically turned into a customer support guy for many months. We shipped a special bugfix update outside of the normal service pack release cycle to try and address the bugs. This was kind of unheard of at the time, and as far as I know we were the first feature team in Office to do something like that.

As things continued many customers and support folks got my name and routed issues directly to me as a way to fast track issues through the normally labyrinthine and bureaucratic escalation process. At one point, a bug was filed in the Office12 bug database that read: “TO THE ATTENTION OF TAILOR BULTER.” From that point on, I assumed a dual identity. Tyler Butler was a PM, but TAILOR BULTER was a customer service MACHINE. TAILOR BULTER was cranking out bug fixes and workarounds as fast as he could type (and convince engineers to type). TAILOR BULTER was glad-handing customers at dev kitchens, helping them build stuff on top of SharePoint and PRIME. I embraced this alter-ego and hammed it up. I even made a fake Time magazine cover with a silhouetted face that carried the headline, “WHO IS TAILOR BULTER?” The subheading, I think, was something stupidly grandiose like, “SharePoint’s secret weapon in the fight against bugs.” I wish I still had the print-out, but I can’t find it anywhere. It hung on my door right next to the “Should I bother Tyler?” flow chart that humorously attempted to get people to leave me alone when my office door was closed. 😊

I’d like to think that I materially improved many customers’ experiences, and I know we made measurable improvements in the feature through the efforts that I drove. But goodness, was it tiring. I burned out spectacularly, and left the SharePoint team in some ways out of desperation because it was clear to me that as long as I was in SharePoint, I’d be both Tyler Butler and TAILOR BULTER. As awesome as TAILOR BULTER was, it was time to leave him behind. As far as I know, he’s still over there in SharePoint-land, kicking the crap out of PRIME bugs. 😉

The fastest-growing WAC application

After leaving SharePoint at the end of Office 14, I moved to a newly created Office team, the Meetings Experience Team (MET). It was there that I met Hani Saliba, who’d been a dev lead on the WAC team previously and became the dev manager for the MET team. It’s also where I met Abraham Mathew, Charlie Burns, Jade Bissat, and Kenneth Mui.

The MET team had a weird charter, and much of the early days there were spent trying to figure out what to build. It was a weird experience for me because there was very little legacy code that we supported. Coming straight out of PRIME land, it was a relief. But one of the drawbacks of the team was that nobody knew what we were building. Not just “other people” - even we didn’t. 😊

Through a variety of windy twisty circumstances, we eventually built (among other things) the O15 version of PPT and Word broadcast. This is when I was introduced to WOPI as well as Nick Simons and Matt Ruhlen. Getting MET plugged into WAC Server was a moment that I’ll always remember, because it was the first time in MET that I felt like I knew what we were building (whether it was the right thing to build or not is a separate question). Things ‘clicked’ for me once we put the two things together.

I’ll also never forget how intimidated I was by Matt and Nick. MET worked closely with the Lync team, and the relationship was fraught at best. I desperately wanted to have at least one good relationship with a partner team, so I focused on being the best WAC partner I could be. I distinctly remember having meetings with Lync, then turning around and having meetings with WAC (mostly Matt and Nick, though sometimes others), and trying to mediate between two teams that were uhhhhhhh… passionate about architecture. Eventually it became clear that we were going to have to get Matt in the room with an engineer from the Lync team and get some direct discussions going on, because MET serving as a go-between between WAC and Lync wasn’t working well.

Charlie and I had a nickname for the Lync dev lead that we worked most closely with: Vlad the Impaler. He was a pretty gruff guy. Not rude or disrespectful, but he was no-nonsense, somewhat humorless, and he’d ask pointed questions in a way that would make you feel about an inch tall. He was intimidating, and my blood pressure rose just seeing his name in my inbox, let alone being in meetings with him. Like I said, the relationship was fraught. 😊 I wasn’t actually at the first meeting with Matt and Vlad the Impaler. Charlie or Abraham were, I think, and Matt obviously was. I don’t know exactly what transpired, but things got easier after that meeting. It wasn’t smooth sailing, but it went better.

I learned two big things from that experience:

  1. Sometimes the best thing you can do is get out of the way and let people duke it out. While I don’t think you should invite conflict as a general rule, conflict can be an important part of change.
  2. Nick Simons and Matt Ruhlen were people I wanted in my corner. This remains true today. 😁

HotWax

While on the MET team, I got a crazy bad case of insomnia. Windows Phone had just been released, so I started working on a phone app during the night because I couldn’t sleep. My then-manager, Nate Stott, made an offhand comment to me that I should consider building something that could be useful at work. We had a need for some perf analysis tools at the time, and that’s what he was thinking about. But after learning about WOPI, I thought to myself, “I could build a WOPI host. That way I’d know WOPI inside and out, and I’d be better equipped to make intelligent design decisions for the MET additions to WOPI. Plus maybe I’ll find some bugs!” Remember earlier when I said I wanted to be a good WAC partner? It was clear to me that Nick and Matt were going to be much more receptive to the changes we wanted to make if they were designed by someone with an intimate knowledge of the protocol. So I took Nick’s draft copy of the MS-WOPI documentation and got to work.

A few weeks later, I showed Charlie a super-secret prototype of viewing documents in Dropbox. I called the app HotWax (because it made the WACs ‘hotter’ lol). He had what I now call “The CharlieB Reaction.” He watched my little demo in silence, nodded once, and asked, “Can you edit docs?” “No, it’s view-only.” “It’s shit. You can’t even edit documents. Shitty.” And then he walked out…

Well, it wasn’t quite that dire, but it was a standard Charlie reaction. UNIMPRESSED. 😊

Fortunately for me, I didn’t take his reaction too seriously, and when the Garage folks advertised a Productivity Science Fair for side projects, I decided to enter HotWax. I built two additional ‘proxies’ for Google docs and Evernote. I built a Greasemonkey plugin for Firefox that would redirect doc clicks in Dropbox through HotWax. It made my demo really amazing because I didn’t have to go through the HotWax site to open the docs in WAC. I made flyers for the fair, put together a pitch deck and an architecture deck to show how it worked (it’s so simple it’s stupid), and best of all, a pretty awesome demo.

During the fair, I met Kyle Ryan and Mark Fields when they stopped by my booth. They both seemed really impressed, and it was exciting to see engineers on the WAC team excited about something I was building during my sleepless nights. Eventually word got around to the WAC team (Kyle and Mark might have had a hand in that), and I did a demo to Joe and others on the team. Once O15 was out the door, the MET team was disbanded, and Mike offered me a PM position on the WAC team. And the rest, as they say, is history. 😊

Of all of the great experiences I’ve had at Microsoft, the one that has been most satisfying to me is the opportunity I’ve had to make the idea of HotWax a reality through the Cloud Storage Partner Program. It’s still amazing to me that a nights and weekends project eventually became a real thing, and I got to be a part of that real thing! Mad props to Nick and Mike for helping make that a reality, and for letting me be a part of it.

Save some for later

I know this is a really long email, but I’ve had a lot of fun reminiscing today and looking through my old OneNote notebooks from my early years. There are more stories I could tell, such as my experiences first meeting Chris Brown and Chris Broussard, the telemetry dream team feature crew with Charlie and AniGo and how we discovered and solved a problem no one at the time knew existed, what it was like being Franklin 2.0 when I joined the WAC team, my first meeting with Andrew Sweeney (with Charlie) when I was still on the MET team, Ben Schmitlin’s scathing email to the Windows team about the Vista Windows Explorer UX, how I managed to piss off annoy an entire team in Office with a blog post… there are more great stories for sure. But it’s best to leave your audience wanting more, so I’ll save the rest for later. Maybe my 15th birthday? For now, this is TAILOR BULTER, signing off. 😊