The man who shipped Windows, kept the Surface a secret, and wrote everything down so you wouldn't have to take his word for it.
Board Partner at a16z. Former President of Windows Division, Microsoft. Author of Hardcore Software - a first-person account of the PC revolution that nobody inside Microsoft was supposed to tell.
There is a version of Steve Sinofsky the industry tells itself. He's the abrasive executive who shipped Windows 7, blew it with Windows 8, and left Microsoft under a cloud. That version is wrong in almost every detail that matters.
The real Sinofsky is something rarer: a person who built mass-market software at a scale almost nobody alive has matched, kept obsessive written records of how he did it, and then spent the next decade teaching the rest of the industry what he learned. He's been writing online since the days of USENET. The Internet you use today exists in part because of an email he sent to Bill Gates in 1994. His engineering blog during Windows 7 produced more pages than most business books and was read by millions of developers. When he finally left Microsoft, he didn't write a tell-all - he wrote a serialized memoir, chapter by chapter, fact-checked against internal documents, that runs to 768 pages and counting.
That quote came years after Windows 8. It's his honest accounting of what happened when he pushed the PC too hard toward touch, too fast, for a market that wasn't ready. That kind of self-examination - public, specific, non-defensive - is unusual in an industry that treats failure as contagious. It's also exactly why he's worth reading.
Today Sinofsky is a Board Partner at Andreessen Horowitz, where he sits on the boards of companies like Everlaw and works with founders building enterprise software. He is not a person who shows up to board meetings and speaks in generalities. He brings 23 years of building software at scale - the kind of scar tissue that can't be faked and can't be taught in a classroom.
His newsletter, Hardcore Software on Substack, is unlike anything else in tech writing. It is not a newsletter. It is a first-person history of the PC revolution told by one of the people who made it. Office 95. Office 97. The Ribbon. Clippy. Windows 7. Windows 8. Surface. All of it, chapter by chapter, with real decisions and real consequences. He publishes it week by week, the way a Victorian novelist published fiction. He has been at it since 2020 and is still going.
The man who nearly ran Microsoft now writes about what it was actually like to do so. The audience is paying attention. So should you.
In January 1994, during a snowstorm, Steven Sinofsky visited his alma mater Cornell. What he found surprised him: students weren't using Microsoft products. They were using web browsers - early, clunky, not-from-Redmond web browsers. He sat down and typed an email to Bill Gates with the subject line: "Cornell Is Wired!"
The email described what he saw: network-connected students browsing a new thing called the World Wide Web, communicating through this distributed system that wasn't Netscape and wasn't Microsoft and wasn't anything Microsoft had thought deeply about. It was a small observation from a mid-level employee on a snowy campus in upstate New York.
It helped trigger one of the most important memos in tech history. Gates' "Internet Tidal Wave" memo of 1995 redirected all of Microsoft toward internet-connected products. It led to Internet Explorer. It led to MSN. It reshaped a $20 billion company around a technology that was barely two years old.
Sinofsky didn't write the Tidal Wave memo. But he planted the seed. And he's been writing things down ever since.
Chapter 2From 1994 to 2006, Sinofsky ran the Office division. It's not glamorous work, building productivity software. Nobody writes profiles about the person who shipped Excel. But Office was - and in many ways still is - Microsoft's most profitable product line. And Sinofsky ran it the way he ran everything: with meticulous, sometimes punishing discipline.
Six major releases. Office 95, 97, 2000, XP, 2003. Each one on schedule. Each one shipped with the features that were ready, not the features that were promised at a conference. He oversaw the integration of Outlook, the development of OneNote, the introduction of SharePoint, and - yes - the Ribbon, the interface redesign so controversial that users actively petitioned to have the menus returned to them.
He also oversaw Clippy. He does not discuss Clippy with nostalgia.
What he built during these years was not just software. It was a model for how to build software: clear accountability, transparent communication, honest assessment of what's ready and what isn't. He published his thinking internally before it was fashionable to blog. He wrote long, detailed posts to his team about decisions, trade-offs, and reasoning. He treated writing as a management tool, not a PR exercise.
Chapter 3When Sinofsky took over the Windows division in 2006, the previous OS - Vista - had just shipped to a reception that was, to use a charitable word, mixed. Enthusiasts called it a disaster. OEMs were furious about driver compatibility. Customers were confused. The press was brutal.
His response to this was not to hire a better PR firm. It was to start a blog.
The "Engineering Windows 7" blog became one of the most ambitious acts of corporate transparency in the history of the software industry. Over the course of Windows 7's development, Sinofsky and his team published over 1,000 pages of detailed, technical posts explaining exactly how they were building the OS - what decisions they were making, why they were making them, what trade-offs they were accepting, what feedback they were incorporating from the public.
It worked. Windows 7 launched in October 2009 to near-universal praise. It reached 450 million users. Microsoft posted record revenue in 2010. Sinofsky was, by any measure, the most powerful product leader at the company and its presumptive next CEO.
Chapter 4While leading the 4,000-person Windows division building Windows 8, Sinofsky was also running a second project entirely hidden from the rest of Microsoft. Hidden, even, from the OEM partners - the Dell's and HP's of the world who built most of Microsoft's hardware ecosystem.
The project was the Microsoft Surface. The first Microsoft-designed, Microsoft-manufactured personal computer. It was so secret that when it was announced in June 2012, OEM partners found out the same way everyone else did: from the news. Relations with those partners - already tense from Windows 8's touch-first design philosophy - got considerably worse.
But the Surface shipped. Both the Surface RT (ARM-based) and Surface Pro (Intel-based) launched in 2012 and 2013. They were imperfect. They were controversial. They were also the beginning of a hardware line that still exists today, refined into some of Microsoft's most acclaimed products.
The model shop team, during Surface development, built a skateboard out of Surface parts. Sinofsky rode it. That is the most delightful data point available about what it felt like to work on that project.
Chapter 5In November 2012, less than one month after Windows 8 launched, Steven Sinofsky left Microsoft. Both sides called it mutual. The circumstances remain, to this day, somewhat opaque. The departure briefly knocked billions off Microsoft's market cap, which tells you something about how the market valued his presence there.
What followed was quieter but arguably more interesting. He joined Andreessen Horowitz as a Board Partner in 2013, one of the more unusual moves in Silicon Valley - a former operating executive transitioning directly into a top-tier VC role without a stint as founder or CEO in between. He brought something rare: not the ability to spot trends, but the ability to run things. To sit on a board and tell a founder, from experience, what actually happens when you decide to hide a product from your OEMs.
He invested early in companies like Tanium, Product Hunt, Everlaw, and Box. He wrote extensively for a16z on topics from AI to enterprise software to the geopolitics of technology. He remained, in other words, exactly what he had always been: a person who thought carefully about software and wrote it all down.
And then, in 2020, he started the newsletter.
A sampling of how industry observers, colleagues, and press have described him over the years.
Sinofsky's reputation at Microsoft was a dual-edged thing. He ran his division like a precision instrument, and other divisions like sharp objects to be avoided. He refused to let other teams coordinate features with his unless it met his standards. He refused to promise features publicly until they were done. He was, by multiple accounts, not easy to work with if you were outside his org.
Inside his org, the picture is more complicated. The 1,000 pages of Windows 7 blog posts were not a PR move - they were a management style. He communicated with his team in writing because he believed writing was thinking. He believed that if you couldn't explain a decision clearly in prose, you hadn't made the decision clearly in your head.
Slate called him "Microsoft's Steve Jobs." The comparison is fair in one respect: he refused consensus. Where Jobs used aesthetic intuition, Sinofsky used product discipline. Both were right more often than they were wrong. Neither was easy to work with when they thought you were wrong.
He also decorated his laptops with stickers. Studied Russian in college. Rode a skateboard made of Surface parts. The full picture is more interesting than the legend.