The pony had other plans. Alex Hinrichs was supposed to ride into a Microsoft team meeting behind executive Terry Myerson, deliver bad news about a slipping Windows Phone milestone, and do it all in a clown suit. The meeting was inside an outdoor circus tent. Myerson, dressed as William Wallace, entered first on a tall white stallion. Hinrichs's pony began trying to throw him. He dismounted, borrowed the calmer horse, rode in to laughter, climbed onstage, and told more than a thousand people that the release candidate was not ready. The crowd groaned, then demanded a dance. Steve Ballmer joined the scene and later peered through the makeup: “Alex? Is that you?”
Yes, Steve. It was Alex. Beneath the greasepaint was one of Microsoft's long-serving specialists in the craft that begins after the grand idea and ends, if fortune and discipline cooperate, with a product leaving the building. Hinrichs was a release manager. He made immense groups of engineers, features, defects, promises, and dates cohere long enough to ship.
The circus story is funny because the costume is absurd. It is useful because the message was not softened. The team had worked hard and still missed the bar. Hinrichs's job was to say so plainly, absorb the disappointment, and direct everyone toward the next attempt. A schedule can survive a clown suit. It cannot survive make-believe.
A five-year plan that lasted twenty-two
Hinrichs arrived on the Microsoft campus in June 1995 as a program-management intern working on Windows NT. He was a Texan in a necktie. His manager was in flip-flops and advised an immediate wardrobe correction. Hinrichs expected to spend five years learning, making friends, and finding people with whom to start a company. Instead, each internal opportunity proved more tempting than departure. His first full-time job began in 1996. Microsoft became home for 22 years.
The product list is a compact history of enterprise and consumer computing: Small Business Server, BackOffice, Windows Server 2008, Windows Phone 7, 8 and 8.1, the first HoloLens, then Windows Mixed Reality. He counted 12 releases in all. The scale climbed, but the essential challenge remained recognizable. Vast human systems had to balance time, features and quality while speaking just enough of the same language to move in one direction.
Windows Server 2008 put the scale in public view. More than 5,000 people contributed over three years. In the room where the final build reached manufacturing, Hinrichs described an almost suspicious calm: the remaining bugs were few, testing had gone smoothly, and the component teams had signed off. Shipping, at its best, looked less like a last-minute rescue than the quiet consequence of many earlier decisions.
His own accounting offers another clue. He wrote more than 250,000 words in weekly messages to the Phone and HoloLens engineering teams. These were not literary side projects. Writing was part of the machinery. His rule was blunt: “Whoever writes it down first wins.” The first coherent version of a decision becomes the thing others revise. In a room full of intelligent disagreement, the author supplies both a map and a pencil.
“Engineering is all about people. Big software projects are built by giant human systems, and giant human systems run on relationships.”Alex Hinrichs, reflecting on 22 years at Microsoft
The mailroom school of decision-making
During the Windows Phone years, the daily ship-room meeting sometimes exposed a decision too urgent to defer and too tangled to settle in the allotted hour. Hinrichs, Myerson, fellow vice president Henry Sanders, and whoever else mattered would cross the hall to the nearest unused room. It happened to be the mailroom. Physical mail was scarce; unresolved software questions were not.
The conversations could become heated. A junior engineer opening the door might find two vice presidents in full debate and retreat at speed. Once the group chose a direction, Hinrichs sat on the mail counter and wrote the decision to the wider team. They were not always right. They were, however, no longer stationary.
Indecision, in Hinrichs's telling, is not prudence. It sends a tax down the organization. People wait. Dependent teams wait behind them. Confidence frays. Most choices are reversible, so he recommends a menu of options, a stated recommendation, and a bias toward action. If the facts change, reverse the choice, tell the people doing the work immediately, and explain the new logic to everyone who relied on the old one.
Before any of that, he insists on a more elementary question: what problem are we solving? Smart people often quarrel over solutions because each has quietly brought a different problem into the room. Hinrichs asks teams to state the problem, test whether it is worth solving, then open the gates to possible answers. It is meeting hygiene with the severity of doctrine.
The first HoloLens, and the miracles department
HoloLens was the release he chose for posterity. He once wrote that he wanted the second sentence of his obituary to read, “Release Manager for the first HoloLens ever created.” His explanation was unusually candid for a product retrospective: the undertaking looked implausible. The physics problems seemed insoluble, the schedule too tight, manufacturing too exacting, thermals too high. Too many miracles were required.
Then the team did it. The sentence matters less as a claim of individual invention than as a release manager's delight in collective improbability. His work lived in dependencies and handoffs, the territories that disappear when a finished device is held up under stage lights. HoloLens supplied the spectacle. Hinrichs helped supply the finish.
He understood failure just as plainly. Windows Phone was, in his judgment, a good product delivered two years too late to win the app platform. He loved it anyway and carried a Lumia 1020 running Windows Phone 8.1 through 2017, long after the market had rendered its opinion. There is something both stubborn and tender in that choice: the release manager living with the released thing.
The people test
After leaving Microsoft in early 2018 and returning to Texas, Hinrichs consulted with Facebook's Oculus product group. During his first week in the halls, he encountered roughly a dozen former Microsoft colleagues. The meetings came with high fives and hugs. They had survived the arguments and deadlines, and they were pleased to be in one another's company again. Hinrichs has said that outcome means nearly as much as anything.
It clarifies his otherwise brisk advice. He favors direct feedback, immediate and private. His preferred structure is “keep, stop, start”: acknowledge what works, identify what does not, and name the new behavior. Ask permission to coach. Put away the phone. Praise effort but critique specifics. Firmness, for him, need not be theater. It is a form of support when delivered with care.
His definition of leadership has three parts: direction, clarity, support. Direction supplies the north star. Clarity marks what is in and what is out. Support means resources, introductions, coaching, removed barriers and air cover. The order is tidy; actual management rarely is. That is precisely why he writes the order down.
“Be direct. Be firm. But don't be a jerk. Instead, be kind.”Alex Hinrichs on the movement he would like to inspire
A second act, still written in verbs
Hinrichs now lives in Houston with, in his own memorable phrasing, a “sparkling wife and a demanding cat.” He serves as a vice president and member of the founding team at Truveta, where large technical systems and large institutional collaborations meet. The subject matter has changed from servers, phones and headsets to a data platform. The operator's problem remains familiar: people with different expertise must agree on what they are building and how they will know when it is ready.
His writing archive, built largely around the lessons that followed his Microsoft retirement note, is a manual for that difficult middle. It avoids the glamour of beginning and the relief of completion. Its territory is the argument before the choice, the uncomfortable feedback after the meeting, the owner missing from the plan, and the deadline that exists only after somebody says it aloud.
One early-career memory supplies the gentlest version of the philosophy. In his first month at Microsoft, Hinrichs attended a training talk by a respected director named Melinda. She mentioned Dallas; he introduced himself as a fellow Texan from Houston. Only afterward did his astonished peers tell him that Melinda French was Bill Gates's wife. His ignorance had spared him the intimidation of status. The lesson he kept was to treat everyone as a person and be willing to speak with anyone.
That may be the line running beneath the product names and enormous counts. Giant systems are composed of ordinary encounters. Someone asks the hard question. Someone admits the date will slip. Someone writes the decision. Someone offers cover while a colleague fixes the problem. Then, occasionally, the impossible object ships - and years later, the people who made it are happy to see one another in the hall.
Continue reading