Gergely Orosz
Author of The Pragmatic Engineer newsletter, former engineering manager at Uber
Where to hear and read about Gergely Orosz1 episode or test
Gergely Orosz is the author of The Pragmatic Engineer newsletter and a former engineering manager at Uber who left big tech to write full time. The most useful lesson is that his overnight success rested on years of unpaid writing, real industry depth, and deadlines he set so he would actually ship.
I fixed it by telling people you're gonna get this every week and now I have to do it
Leaving big tech to build the #1 technology newsletter | Gergely Orosz (The Pragmatic Engineer)
Gergely Orosz: the story and the lessons
1Gergely Orosz's path from Hungary to Uber to writing
Orosz climbed through faster-moving companies one step at a time. He did a five-year computer science degree in Hungary and worked at a small web agency on the side. He then worked at a UK consulting company, a London investment bank, Skype, Skyscanner and Uber in Amsterdam, rising from senior engineer to manager to manager of managers.
He was on track for senior engineering manager and maybe director, and left without planning to write. He says climbing slowly gave him appreciation for every level of the job.
A 2016 promise to himself set up his later risk-taking. At Skyscanner he was a principal engineer on about £90,000 base. Uber doubled his pay after he negotiated, mostly in stock, and he promised himself that if Uber went public he would take a risk.
He saw the stock as a lottery ticket and had only about six months of savings. Four years later the pandemic layoffs cut half of his 30-person group.
The Uber layoffs broke his trust and made him want control. People on his team were let go and he was not told who. He felt like a number, and colleagues he rated higher were cut because they sat in the wrong team.
As a middle manager he had no budget and could only complain to HR when someone was underpaid. He decided to leave after a two-week July holiday and served a year of notice, since Europe has longer notice periods.
Self-published books turned a startup plan into a newsletter trial. He planned to raise venture money for a platform engineering tool. Instead his self-published books earned about $100,000 in a year, and he kept about 90% of the revenue.
Daily book income swung between about $10 and $300, so he wanted recurring revenue. He gave the newsletter a six-month trial, told his wife, and promised refunds to annual subscribers if it failed.
2How Orosz built The Pragmatic Engineer
Six years of blogging sat behind the launch. He launched with about 10,000 Twitter followers, got 100 sign-ups on day one and reached 1,000 paid subscribers in six weeks. He had blogged for years and held jobs at Skype, Skyscanner and Uber.
He says it "sounds like a fairy tale" but calls it "at least six years of kind of accidental work behind this". Readers who knew the blog signed up for a weekly email.
Orosz doubled down when a draft drew unusual demand. A long draft on mobile engineering at scale brought 20 messages in an hour, against his usual three or four a day. He made it a pay-what-you-want ebook and spent two months on it.
The book was free for those two months and got a sponsor. He never expected people to care about large mobile apps, but thousands did.
He writes only on topics paid readers cannot find elsewhere. Each post takes about a week of research, drafting, feedback and editing, and often runs five to six thousand words. Tuesday is the in-depth, timeless post and Thursday is a timelier scoop.
He talks with people to research, gets trusted readers to review, and works with an editor who is a journalist. His reasoning: "why would people pay for something that is out there already or well known".
He tracked every article's effect on annual revenue. For the first year he logged in a spreadsheet how much his annual revenue rose a week after each post. Good posts acted like a raise.
Sometimes a post he thought mediocre did well, so the log gave fast feedback on what readers valued.
Substack's recommendations feature drove the growth burst. The first nine months brought 50,000 subscribers, and the next six or seven months added more than 100,000. He now has about 189,000, with a single-digit percentage paying.
Revenue comes only from subscriptions, with no ads or sponsorships. He earns more than his peak Uber pay of roughly $320,000 to $330,000, and says part of that is luck and situational.
3Orosz's rules for working alone
He builds outside deadlines because self-discipline vanished. After almost 15 years as a diligent employee, his discipline disappeared once he worked for himself. He promised readers a post every week, later two, so pressure hits almost every day.
He says to sign with a publisher or find someone to hold you accountable. A book offer of about $5,000 taught him that the value of a publisher is the pressure, not the money.
A 20-minute timer plus a site blocker gets him started. When he cannot start, he sets a 20-minute timer and runs a Python script that edits his hosts file to block Twitter, LinkedIn and Facebook. After about five minutes of grumbling he is heads down.
He has to rerun the script to unblock, which adds friction. Chrome extensions can do the same for non-technical people.
He rotates productivity tools because each stops working. Pomodoro with 25-minute intervals and the Centered app each helped for a few months before his brain learned to work around them.
Centered's video camera feature worked because it felt like someone was watching. He prefers his own scripts to tools that feel too opinionated or not opinionated enough.
He protects an empty calendar and plans to cut his hours. At Uber his calendar was meetings on top of meetings. Now he has almost no meetings and long stretches of deep work.
He wants to go from about 50 hours a week to 20 to leave room for the next idea. He has stopped making long-term plans, noting that three years before he aimed to be a manager of managers and never pictured a newsletter.
Solo work costs him a team, certainty and an easy exit. He misses the people at Uber and rents a shared office for techies. Newsletter peers each run their own business, so they do not share everything. Without meetings he felt guilty and unproductive for months.
Annual plans mean readers expect something good every week for a year, and a newsletter cannot be sold for what its revenue suggests. He calls his own success an outlier.
He now plans time off and works ahead on lighter pieces. He took no holiday in his first year, which caused friction with his family. He now works ahead on less time-sensitive posts and takes about four weeks off a year.
At Uber he took a four-month paid leave and logged off completely. He says of the solo business: "it's hard to turn off".
4Orosz's advice on starting to write
Orosz says to become an expert before teaching. He chose well-known employers on purpose, coming from a small country and a little-known university. He had blogged and spoken at conferences before Uber.
He notes Kent Beck kept working inside the industry, joining Facebook at a lower title. His view: "become an expert somewhere somehow before you start because you'll be a lot more credible".
He sets goals he controls, not outcomes. As a manager he called promotion and 20,000 subscribers bad goals because they are out of your control. Good goals are learning a language this year or leaving work at 5 pm on Fridays.
His first blog goal was one post a month, and he was proud of keeping it. Step one is to start, step two is to keep it up.
He tried Jeff Atwood's steady-posting formula and failed first. Atwood's advice was a post three times a week for two years. Orosz tried one every two weeks, wrote six and gave up.
Later a post arguing that a code comment is an invitation to refactor spiked on Hacker News and crashed his shared hosting. Seeing engineers argue over his writing showed him people cared.
He runs time-boxed experiments, then decides. He gave the newsletter a six-month window. He cites Steve Yegge, who committed to six months of a weekly podcast, saw a few thousand subscribers, stopped and moved on.
His four habits for success are depth, a format, a cadence and experiments. In his words: "start something do a Cadence see if it sticks".
He picked blogging because depth is less crowded there. A YouTuber with 200,000 subscribers told him few blogs publish in-depth work regularly. Video brings a bigger audience but far more competition.
He suggests deciding between entertainment, education or edutainment, then choosing a format that fits you. He prefers writing for engineers because they can scan it, and he enjoys it enough that it feels like fun.
He tells people to start side projects and share what they know. Nobody at work appreciates a newsletter, so he says to start small on the side through blogs, meetups, talks or video.
He once wanted only to climb the corporate ladder with writing on the side, and now it has flipped. He still calls much of his result luck.
Leaving big tech to build the #1 technology newsletter | Gergely Orosz (The Pragmatic Engineer)
Orosz explains how six years of consistent blogging let him launch a paid engineering newsletter that hit 1,000 paid subscribers in six weeks, and how he structures a solo creator business. He now earns more than his roughly $320-330K peak compensation at Uber, with about 189,000 subscribers.
Orosz climbed through faster-moving companies one step at a time. He did a five-year computer science degree in Hungary and worked at a small web agency on the side. He then worked at a UK consulting company, a London investment bank, Skype, Skyscanner and Uber in Amsterdam, rising from senior engineer to manager to manager of managers.
He was on track for senior engineering manager and maybe director, and left without planning to write. He says climbing slowly gave him appreciation for every level of the job.
A 2016 promise to himself set up his later risk-taking. At Skyscanner he was a principal engineer on about £90,000 base. Uber doubled his pay after he negotiated, mostly in stock, and he promised himself that if Uber went public he would take a risk.
He saw the stock as a lottery ticket and had only about six months of savings. Four years later the pandemic layoffs cut half of his 30-person group.
The Uber layoffs broke his trust and made him want control. People on his team were let go and he was not told who. He felt like a number, and colleagues he rated higher were cut because they sat in the wrong team.
As a middle manager he had no budget and could only complain to HR when someone was underpaid. He decided to leave after a two-week July holiday and served a year of notice, since Europe has longer notice periods.
Self-published books turned a startup plan into a newsletter trial. He planned to raise venture money for a platform engineering tool. Instead his self-published books earned about $100,000 in a year, and he kept about 90% of the revenue.
Daily book income swung between about $10 and $300, so he wanted recurring revenue. He gave the newsletter a six-month trial, told his wife, and promised refunds to annual subscribers if it failed.
Six years of blogging sat behind the launch. He launched with about 10,000 Twitter followers, got 100 sign-ups on day one and reached 1,000 paid subscribers in six weeks. He had blogged for years and held jobs at Skype, Skyscanner and Uber.
He says it "sounds like a fairy tale" but calls it "at least six years of kind of accidental work behind this". Readers who knew the blog signed up for a weekly email.
Orosz doubled down when a draft drew unusual demand. A long draft on mobile engineering at scale brought 20 messages in an hour, against his usual three or four a day. He made it a pay-what-you-want ebook and spent two months on it.
The book was free for those two months and got a sponsor. He never expected people to care about large mobile apps, but thousands did.
He writes only on topics paid readers cannot find elsewhere. Each post takes about a week of research, drafting, feedback and editing, and often runs five to six thousand words. Tuesday is the in-depth, timeless post and Thursday is a timelier scoop.
He talks with people to research, gets trusted readers to review, and works with an editor who is a journalist. His reasoning: "why would people pay for something that is out there already or well known".
He tracked every article's effect on annual revenue. For the first year he logged in a spreadsheet how much his annual revenue rose a week after each post. Good posts acted like a raise.
Sometimes a post he thought mediocre did well, so the log gave fast feedback on what readers valued.
Substack's recommendations feature drove the growth burst. The first nine months brought 50,000 subscribers, and the next six or seven months added more than 100,000. He now has about 189,000, with a single-digit percentage paying.
Revenue comes only from subscriptions, with no ads or sponsorships. He earns more than his peak Uber pay of roughly $320,000 to $330,000, and says part of that is luck and situational.
He builds outside deadlines because self-discipline vanished. After almost 15 years as a diligent employee, his discipline disappeared once he worked for himself. He promised readers a post every week, later two, so pressure hits almost every day.
He says to sign with a publisher or find someone to hold you accountable. A book offer of about $5,000 taught him that the value of a publisher is the pressure, not the money.
A 20-minute timer plus a site blocker gets him started. When he cannot start, he sets a 20-minute timer and runs a Python script that edits his hosts file to block Twitter, LinkedIn and Facebook. After about five minutes of grumbling he is heads down.
He has to rerun the script to unblock, which adds friction. Chrome extensions can do the same for non-technical people.
He rotates productivity tools because each stops working. Pomodoro with 25-minute intervals and the Centered app each helped for a few months before his brain learned to work around them.
Centered's video camera feature worked because it felt like someone was watching. He prefers his own scripts to tools that feel too opinionated or not opinionated enough.
He protects an empty calendar and plans to cut his hours. At Uber his calendar was meetings on top of meetings. Now he has almost no meetings and long stretches of deep work.
He wants to go from about 50 hours a week to 20 to leave room for the next idea. He has stopped making long-term plans, noting that three years before he aimed to be a manager of managers and never pictured a newsletter.
Solo work costs him a team, certainty and an easy exit. He misses the people at Uber and rents a shared office for techies. Newsletter peers each run their own business, so they do not share everything. Without meetings he felt guilty and unproductive for months.
Annual plans mean readers expect something good every week for a year, and a newsletter cannot be sold for what its revenue suggests. He calls his own success an outlier.
He now plans time off and works ahead on lighter pieces. He took no holiday in his first year, which caused friction with his family. He now works ahead on less time-sensitive posts and takes about four weeks off a year.
At Uber he took a four-month paid leave and logged off completely. He says of the solo business: "it's hard to turn off".
Orosz says to become an expert before teaching. He chose well-known employers on purpose, coming from a small country and a little-known university. He had blogged and spoken at conferences before Uber.
He notes Kent Beck kept working inside the industry, joining Facebook at a lower title. His view: "become an expert somewhere somehow before you start because you'll be a lot more credible".
He sets goals he controls, not outcomes. As a manager he called promotion and 20,000 subscribers bad goals because they are out of your control. Good goals are learning a language this year or leaving work at 5 pm on Fridays.
His first blog goal was one post a month, and he was proud of keeping it. Step one is to start, step two is to keep it up.
He tried Jeff Atwood's steady-posting formula and failed first. Atwood's advice was a post three times a week for two years. Orosz tried one every two weeks, wrote six and gave up.
Later a post arguing that a code comment is an invitation to refactor spiked on Hacker News and crashed his shared hosting. Seeing engineers argue over his writing showed him people cared.
He runs time-boxed experiments, then decides. He gave the newsletter a six-month window. He cites Steve Yegge, who committed to six months of a weekly podcast, saw a few thousand subscribers, stopped and moved on.
His four habits for success are depth, a format, a cadence and experiments. In his words: "start something do a Cadence see if it sticks".
He picked blogging because depth is less crowded there. A YouTuber with 200,000 subscribers told him few blogs publish in-depth work regularly. Video brings a bigger audience but far more competition.
He suggests deciding between entertainment, education or edutainment, then choosing a format that fits you. He prefers writing for engineers because they can scan it, and he enjoys it enough that it feels like fun.
He tells people to start side projects and share what they know. Nobody at work appreciates a newsletter, so he says to start small on the side through blogs, meetups, talks or video.
He once wanted only to climb the corporate ladder with writing on the side, and now it has flipped. He still calls much of his result luck.