Leaving Uber to Write Full Time: Gergely Orosz on Building The Pragmatic Engineer

Orosz describes his early career as a fairly typical engineering path. He grew up in Hungary, studied computer science, moved to the UK, and worked his way through consulting, an investment bank in London, Skype, Skyscanner and finally Uber in Amsterdam, where he went from senior engineer to manager to manager of managers.

Just as he seemed headed for senior engineering management, and maybe a director role, he decided to leave. He did not plan to write a newsletter, but that is where he ended up, and it has turned out to be what he calls a really cool life.

The numbers are striking. He reports about 189,000 subscribers, with roughly 80,000 added in the previous 90 days, and growth accelerated after Substack introduced recommendations. A single-digit percentage of readers pay, which still means thousands of paying subscribers, and he earns mostly from subscriptions rather than sponsorships or ads.

In his best year at Uber, after the IPO, Orosz made about $320-330K in total compensation, which he says is very high for Europe and roughly equal to $500-600K for the same role in the US. When he quit he assumed he would earn far less in exchange for enjoying the work. Instead he now makes more than at Uber.

The difference is structural. A newsletter's earnings rise as long as it grows, subject to churn and slowing growth. For the first year he kept a spreadsheet of how much his annual revenue rose a week after each article, and a standout piece effectively gave him a raise, something he says corporate life does not offer.

Host Lenny Rachitsky adds his own framing: instead of one boss who can fire you, you have thousands of micro-bosses, and some "fire" you daily by unsubscribing. He argues that still feels safer than a job where one person can decide your fate. Orosz also attributes part of his result to luck and circumstance.

When Uber hired him in 2016 it roughly doubled his compensation, most of it in stock, and asked him to build a SOX-compliant payments team, a hint that an IPO was coming. He told himself that if Uber exited and he had four years of savings, he owed himself a risk. Four years later, almost on the dot, COVID hit and he had to lay off a big share of his group of about 30.

His plan was to leave, finish his *Software Engineer's Guidebook*, and then raise venture capital for a platform-engineering startup, a path many Uber alumni take. He got sidetracked: he accidentally published *Building Mobile Apps at Scale* and a book on tech resumes, which earned about $100,000 in their first year through self-publishing.

Before committing to a startup he asked himself why he wanted it. He missed small, scrappy teams and did not enjoy managing managers, but when he imagined the end state, a large payout followed by writing and teaching, he saw the fisherman parable. A businessman tells a fisherman to work for years so he can eventually fish for three hours a day, which is what the fisherman already does.

Books made income hard to predict, so he chose a paid newsletter for its recurring revenue. After collecting a long list of ideas he announced it, told his wife he would give it six months like a startup, and planned refunds if it failed. The layoffs, he says, shattered his trust that the company would take care of him, and he had also found middle management frustrating because he lacked real authority or budget.

Orosz started with one in-depth post a week and now publishes two: a timeless engineering deep dive on Tuesdays and a more timely column, the scoop, on Thursdays. A single deep dive takes days to weeks of research, a draft, feedback from trusted readers and an editor, and his emails often run five or six thousand words. Even with a cleared calendar, the early months were stressful.

His weekly cadence is built so that almost every day carries a deadline. Monday he finishes the Tuesday piece, Tuesday he publishes and free-writes, Wednesday is the light day, Thursday is the scoop, and Friday he starts the next Tuesday article. He says that pressure, plus knowing thousands of people pay him, explains how he wrote the equivalent of four or five books last year.

He also admits he misses the structure of his Uber schedule, and recommends external accountability. If you want to write a book, he suggests signing a publisher contract for the pressure rather than the money. His own sneaky plan was to assemble the *Guidebook* from newsletter posts, inspired by Alexandre Dumas writing *The Three Musketeers* in installments for a magazine.

Unstructured time is the problem both newsletter writers face. Orosz notes that Twitter is both a research tool and a promotion channel, so it justifies far more attention than it should. He finds that any method works for a few months before his brain learns to route around it.

He has tried apps such as Centered, whose focus sessions and optional video camera made him feel watched, and the Pomodoro technique. What has consistently worked is a deadline, like needing to finish before going home, and a 20-minute timer combined with a small Python script that edits his hosts file to block LinkedIn, Twitter and Facebook.

The hard part is starting: he grumbles for the first few minutes, then around five minutes in he switches into focus. He also wonders whether his frequent guilt about not working harder is real or just his mind asking for a break. Lenny sums up the three most reliable tricks as deadlines, site blocking and a short timer.

What Orosz loves most is an empty calendar after years of meetings piled on meetings, and the creative freedom to change formats and experiment. Lenny agrees and keeps a rule of no meetings before 3 PM to protect deep work.

The downsides are real. He finds it lonely, since he liked chatting at the coffee station at Uber, so he works from a shared tech office. For the first months he felt guilty that he was less productive than when he had eight meetings a day. The work is also stressful in a way he did not expect, with a constant need to judge how he is doing by subscriber numbers and doubts about whether he made a mistake.

Both discuss the lack of an exit. A newsletter is tied to its author, so it cannot easily be sold at full value, unless you build an organization around it. Lenny adds that you must produce something good every week, which makes long breaks risky, and Orosz notes he took no holiday in the first year, which caused friction at home. Both are trying to work ahead so they can take real time off.

Orosz has stopped making long-term plans, noting that he never predicted he would become a manager of managers or a newsletter author. He now treats the newsletter as a one-person business rather than a creator career, which helps him detach and enjoy weekends. He is open to the format changing over time.

One goal is to cut his roughly 50 hours a week on the newsletter to about 20 and leave room for the next spark of inspiration. He points out that the newsletter itself only came because he gave himself six months without work obligations. Lenny warns that any freed-up hours tend to fill with new projects.

The most valuable thing about being in charge of your time, Orosz says, is that you can double down on pull. A long mobile-engineering blog post drew 20 messages in an hour, when he normally got three or four a day, so he turned it into an ebook and then into a sponsored book. The same pattern led him to the newsletter.

The story people tell is that he announced the newsletter to about 10,000 Twitter followers and 1,000 LinkedIn followers, got 100 subscribers on day one and 1,000 paid subscribers within six weeks. Orosz stresses that it hides at least six years of accidental work. He had blogged since graduating, then rebooted the blog as The Pragmatic Engineer after reading Jeff Atwood's advice on getting famous online: write a post, three times a week, for two years.

He lasted six posts before giving up, but a traffic spike from Hacker News crashed his shared hosting. The post, "A comment is an invitation for refactoring", split readers into camps calling him an idiot or a genius. That showed him people might read what he wrote, and he kept posting every few months for years.

By the time he launched the newsletter, he had popular posts on performance reviews and on the trimodal nature of software engineering salaries. Readers who already knew the blog signed up to get the writing weekly. His advice is to get started and keep it up, ideally for yourself, since he never thought it would become a business.

Asked whether you need his background, Orosz points out that he blogged and spoke at meetups before his big-name jobs. His advice is to start teaching and sharing what you know in any format, whether newsletter, video or meetup talk. He suggests setting goals within your control, such as writing monthly, instead of outcomes like a subscriber count.

He also credits pedigree: he deliberately worked his way up to well-known employers, and says people would pay far less attention if he had worked at a company nobody had heard of. Lenny adds that credibility comes from real experience on things that scaled. Orosz uses Kent Beck as a model of someone who keeps working inside the industry, even taking a title cut to join Facebook, so he has fresh stories to share.

Boiled down: become an expert somewhere, share in whatever format suits you, keep a cadence, and be willing to experiment. He cites Steve Yegge, who ran a podcast for six months to see if it stuck and then returned to industry. Orosz also notes that fewer people now write in-depth blogs, because many moved to YouTube and TikTok, which can be an advantage, and he urges choosing between entertainment, education or edutainment.

He closes by inviting listeners to subscribe at pragmaticengineer.com, to send anonymous tips for his Thursday scoop, and to get in touch if they work at Google, since his next article will cover its engineering culture. Lenny sums up the whole conversation as: build depth, then write twice a week for two years.