Disruption: a front-row seat

July 19, 2018

My first time living outside of Illinois was in 2005, as a Microsoft intern. I was sitting there, as a college student in 2005, trying my damndest to understand "current sheets", while the beer-swilling apes I lived with played "banana baseball" at 2AM. I took a full-time job at Microsoft after that internship, in 2009. In retrospect, this whole story feels inevitable; that internship was the first time I was with my people, the nerds, out in Seattle; of course I was headed back after college.

I started full-time in August 2009. I look back on this job fondly for the people I met—guys like Omar—economic migrants from all corners of the world [1], gathered together to invent the future. It was awesome.

This isn't a happy story, though; the people were the redeeming grace of a chaotic, awful job. And while I could've handled it better—I was young, and arrogant—I was nothing but a speck of dust in a tornado that's still tearing through the industry, more than 10 years later.

For a long time, Microsoft was an enterprise technology company; it made money selling software to businesses. Everybody has heard of Windows and Office, but even SQL Server—a product with less name recognition—was an independent billion-dollar/quarter business by 2009. SQL Server was my home at Microsoft.

From the day I arrived, I had little idea what I was doing. Nobody does at their first job, but my manager seemed, to put it nicely, elsewhere. She took little interest in my work, spending her time in meeting after meeting with some guy named Bob Muglia, trying to convince him to fund her vision, that the group's strategy was sound. What little time she spent with my direct group, I spent as her secretary, writing page after page of notes chock-full of some bizarre foreign language. I already had a bit of experience developing software and yet all of it seemed so foreign, words like "stovepipe system", "repository", and "illities".

What the hell were these people talking about?

Throughout my time at Microsoft, I felt like my colleagues lived on some strange island, cut off from the rest of the world. I was reading about Heroku, AWS, and Postgres on Hacker News; they'd never even heard of any of these things, let alone used them. At first I wasn't sure whether I just had a lot to learn, but as the weeks wore on, it dawned on me that they really were clueless. Having spent years in the comfortable ivory tower of Redmond, a chasm had opened between my group's understanding of customer needs, versus the on-the-ground reality. My inquiries when my manager had last got out of the building, were met with hemming, hawing, and nothing concrete. I'm increasingly convinced this is how so many tech giants die: they lose touch, becoming so inwardly focused on their own politics, processes, and turf battles, that they lose sight of what's important: customers, and their needs. [2] Had any of my colleagues opened an IDE, or spent a month with a midmarket IT manager, I think their perspective would've been different—more accurate. They thought they still "made their own gravity", even though by 2009, it was clear the clock had run out.

Only later did it become clear what I'd stepped into: a fight, at the highest levels of the company, over the future of SQL Server. SQL Server had always been positioned as a low-cost alternative to expensive databases like Oracle, DB2, etc., but Postgres and mySQL on the public cloud were starting to rack up "design wins", as Microsoft called them. It was textbook low-end disruption: SQL Server was licensed for $10-20k/processor, and the thought of putting a crown jewel—SQL Server—on offer through Azure for mere pennies/hour to be price-competitive with Heroku Postgres or RDS, was unthinkable. It turned their stomachs. There was just no way a decision that would significantly cut the performance of Microsoft's #3 revenue product was going to get green-lighted, though it was increasingly clear that doing nothing would hasten their slide into irrelevance.

I left after six months. In the months after my departure, the chaos in my department spilled into the open, its full magnitude becoming apparent: my project, "Oslo", was "obliterated", 24-year Microsoft veteran Bob Muglia's career at Microsoft met its end, and some no-name nerd named Satya Nadella was promoted to head of Server and Tools [3]. I don't know this for sure, but I suspect Bob Muglia just couldn't let go of expensive, on-premise software. Many people can't. Bob is CEO of Snowflake Computing now, another enterprise database company, but apparently it's "Build in the Cloud".

Maybe he came around? Hard to say. But the world changes, and if you're not careful, you'll get left behind.

[1] Whenever people take stock of national origin—"do we have people here from every continent?"—Omar is always the token African. He loves seeing people's shocked reactions, because his skin is very light, being from Morocco.

[2] If Google's messaging strategy is any indication, they've started to suffer from this problem in a big way.

[3] At the risk of stating the obvious, this "nerd" later became CEO, a job only two others have held in the company's 40-year history.