This publication has turned three years old, which feels like the right time to pause and reflect on the past 12 months for The Pragmatic Engineer.

At time of publication, 759,402 readers subscribe to this newsletter. This is 300,000 readers up on a year ago, when it was just under 461,000. This figure far exceeds the size of the audience I assumed would be interested in software engineering deep dives into some fascinating and challenging topics, when I launched this publication. Thank you for your trust!

I’d like to extend a very special thanks to all paying subscribers: the publication runs on your support, and it enables us to deliver well-researched deep dives. Many paying subscribers expense this newsletter from their company learning and development budget. If you have such a budget, here’s an email to send to your manager. There’s also reduced prices for people in countries with lower average income than the US and Western Europe, and student discounts. If you enjoy the newsletter, you can also gift a subscription to others.

Speaking of deep dives: over the past year full subscribers received two articles almost every week: a deepdive on Tuesdays, and tech news in The Pulse on Thursdays. The last 12 months of issues adds up to around 5-7 books’ worth of reading (about 550,000 words). Free subscribers also got well over a nonfiction book’s worth of reading in their inbox: the first parts of the Tuesday articles, and a full article monthly.

Today, we look back, and peer in to the future:

1. How The Pragmatic Engineer has evolved

2. Popular, interesting topics

3. What I learned about readers!

4. Notable events

5. Looking ahead, and The Pragmatic Engineer Podcast

Before we jump in: if you’re someone who enjoys podcasts, I have a treat for you. The Pragmatic Engineer Podcast is launching with the first episode next week. To get the first episode in your podcast player, subscribe on:

• YouTube — you can also watch the podcast announcement here

• Spotify

• Apple Podcasts

• … or your favorite player!

1. How The Pragmatic Engineer evolved in the last year

Software engineering is an ever-changing field where the best teams continuously experiment with new approaches, and change how they work. I’m also an engineer at heart, so it’s only natural to do the same with the newsletter. 

Here’s what’s changed in recent months; some of which you might have noticed, and other things that happened behind the scenes.

Unshared details from interesting tech companies

Some companies attract media attention due to how well they execute, and what this publication tries to do is learn how exactly they achieve it. I aim to bring details straight from the source, talking exclusively with engineering teams at companies in the spotlight. The last 12 months has seen exclusive articles on:

• OpenAI: How they scaled ChatGPT, and Inside OpenAI: how does ChatGPT ship so quickly?

• Anthropic: How the company built Artifacts (and details on how it operates)

• Stripe: Inside Stripe’s engineering culture

• Meta: How the company built Threads with a small team in 6 months

• Bluesky: How a tiny team built a Twitter/X competitor and Inside Bluesky’s engineering culture

• Oxide: A hardware/software startup on “hard mode”

In terms of deep dives, it’s been the most prolific year yet for this publication. Doing lengthy articles which lift the lid on how leading teams get things done, can be challenging. Understandably, some companies do not normally share the details you’ve been able to read in this newsletter. At the heart of delivering articles like those above, is trust. Engineering leaders and software engineers ultimately feel they have scope to be candid with me, and that means a lot. Their trust in myself and The Pragmatic Engineer team is a very rewarding part of writing this newsletter.

More articles mixing deep research with pragmatic observations

Since February, it’s not just me doing the extensive research that goes into each article: Elin Nilsson is this publication’s first Tech Industry Researcher. Previously, Elin spent 7 years at Spotify, working in various mobile and platform engineering teams. Before, she interned at both Spotify as an Android developer and Google as a web/UX prototyper. She’s brought a whole lot of things onboard since she joined in February: like her infectious enthusiasm for nerding out about tech topics, diving deep into technologies and trends, and bringing order to the chaos of having too much information to process on any given topic.

Since she’s come onboard, there’s been more deeply researched pieces published than ever, and each one of them went into more depth. Some of the articles that started with Elin’s insightful research:

• GenZ software engineers according to older colleagues, and what GenZ devs really think

• Bluesky: How it was built, and Inside Bluesky’s engineering culture

• How do AI software engineering agents really work?

• AI tooling for software engineers: Reality check, Reality check (part 2) and Rolling LLM tools out company-wide (Part 3)

Thanks to Elin, we can take on ambitious projects that were simply not possible when the Pragmatic Engineer launched because the research can take months, We’re on a roll with Elin, so expect these to keep coming! You can always suggest interesting topics for us to potentially look into. 

Goodbye “Scoop,” hello “Pulse”

During the first two years of this newsletter, there was usually an article every Thursday called “The Scoop,” which rounded up interesting things in the industry, and occasionally broke big tech stories. But as of a year ago, I am no longer doing this and the name change to “The Pulse” was part of this shift. I shared details about the change:

‘The Pulse’ just better represents the mission of these articles, which is to help you keep an “ear to the ground” on what’s happening across Big Tech and at startups – sectors which regularly affect the whole tech industry.

I receive and validate plenty of interesting information from insiders at these companies, but my main focus is analyzing what’s going on in our industry; the implications of that and the opportunities for software engineers and tech businesses.

Names matter, and I feel “The Scoop” suggests a more gossipy, tabloid newspaper-style approach than what these articles actually deliver. Since The Scoop’s first issue I’ve focused on relevant industry changes and what they mean. I hope “The Pulse” makes the intent of this series clear: to provide an ear to the ground, and the latest analysis of the state of our industry.

This change has been one I’m very happy with; it’s also partly why we’ve been able to publish more company deep dives straight from tech companies themselves.

When writing “The Scoop”, tech companies and devs whom I met were never quite certain if I had my “journalist” hat on (and would leak what they shared as news), or my “analytical” software engineer’s hat. 

This made people reluctant to share information that could potentially identify their company, or affect their career. This was taxing for me personally and I decided I don’t want to be in the “gossip” business, but in the software engineering one. For that reason, the change from The Scoop was made.

But while writing it, I learned a lot about journalism and met a lot of reporters. It’s a fascinating world, but one I don’t feel is for me. Also, the time I save on no longer verifying information for  news articles, means there’s more time to dive deep into relevant, interesting software engineering topics.

Industry Pulse

In the spirit of experimentation, I tried out a new section inside the Thursday “The Pulse” article. Industry Pulse is a roundup of tech events relevant for devs and EMs, with commentary.

It has delivered early insights on topics like The impact of Section 174 on the software industry, predicting pressure on commercial open source to make more money, the unfortunate “startup purge event,” analyzing why Amazon sunset several of its products in one fell swoop, and much more.

Readers feedback was very positive from the start. Also, I enjoy writing it and look forward to it every day: so it has stayed!

2. Popular, interesting topics

In case you’re looking for articles to read or re-read, here’s a selection of some standout ones from the past 12 months, in my opinion.

Popular topics

The end of 0% interest rates, and what this means for software engineers was the single most-read article of the past year. This article was in a highly-detailed 4-part series about what feels like the biggest industry change in years: the end of low interest rates.

It is rare for an economic change to have the kind of impact on an industry that higher interest rates are having. We’re finding out that tech startups and interest rates are more connected than many of us would like to think.

Measuring developer productivity? A response to McKinsey. Consultancy giant McKinsey is known for many things, but software engineering expertise is not one of them. Yet the globally known consultancy started selling advisory services about this: including the claim that they had developed a way to measure devs’ productivity.

Dev productivity is a notoriously tricky thing to accurately measure, HEREso I teamed up with industry veteran Kent Beck for a grounded take on why measuring productivity is difficult, and sharing our suggestions on how to go about this kind of measurement.

Inside OpenAI: How does ChatGPT Ship So Quickly? Up to early 2024, OpenAI was dominating headlines by releasing new products and features with an incredible pace. The company was out-executing the likes of Google, Meta and every other AI startup. But how did they do it?

I sat down with Evan Morikawa, who headed up the Applied engineering team (the team also building ChatGPT). We learned how OpenAI operates ChatGPT like a “startup inside a startup,” engineering has an unusually tight integration with Research, has a high talent density in their San Francisco offices, and has a lot of smaller day-to-day habits that add up. While I advise to never blindly copy engineering practices: it’s helpful to understand what makes a startup like OpenAI execute so quickly.

Surprise uptick in software engineering recruitment. Recruitment activity tends to be quiet in the summer months. This is why I was surprised to hear from several developers that they are seeing a spike in LinkedIn reachouts – mostly from startups.

Talking with dozens of developers – and some hiring managers – we uncovered that startups seem to have increased their hiring pace for a variety of reasons. What was common is how they were all searching for experienced engineers, and were becoming more “bold” in their reachouts: pinging senior+ engineers working at Big Tech and well-funded startups, who are profiles that usually are more reluctant to move.

Other interesting topics

Here are articles that I especially enjoyed for one reason or the other. If you’ve not yet read them, they could be worth your time:

• The past and future of modern backend practices. From the birth of the internet, through SOA and virtualization, to microservices, modular monoliths and beyond.

• How Big Tech does Qualty Assurance (QA). Most Big Tech companies have no dedicated SDET, QA, or tester roles. How do they produce quality software? A look into how Microsoft, Google, Meta, Apple, Amazon, Uber and Netflix do it.

• Stacked diffs (and why you should know about them). Meta and Google have been using stacking for closer to a decade: a coding workflow that is very efficient for small PRs. So what is stacking, and how come it’s not more widespread in the industry?

• What is Secruity Engineering? A deep dive into the ever-changing field of security engineering; a domain that can feel intimidating to some software engineers.

• Applied AI Software Engineering: RAG. Retrieval-Augmented Generation (RAG) is a common building block of AI software engineering. A deep dive into what it is, its limitations, and some alternative use cases.

• Thriving as a Founding Engineer: Lessons from the Trenches. Being a founding engineer at an early-stage startup is a vastly different, broader role than many people think.

• Adobe can’t buy Figma: the impact on the industry. Months after we covered Figma’s engineering culture, regulators blocked Adobe’s acquisition of Figma. This event was likely what prompted Big Tech to stop acquiring companies, and to do “talent raids” instead with AI companies they seek to purchase.

There’s more, of course! For the full list of published articles, browse past deepdives, or The Pulse issues.

3. What I learned about readers — about you!

This year was the first time I asked details about who you are: in the form of a survey. 

Disciplines

I write the newsletter for software engineers and engineering managers, and was curious to know if most readers are, indeed, in this field. Turns out this is broadly the case:

The majority of readers are in the software engineering field – ranging from entry-level to senior/lead/staff and above roles – and a good portion is in engineering leadership/management (ranging from engineering managers to VP and C-level engineering leadership positions.)

I’m pleased to hear so many folks in product management following along – extrapolating 5% to the whole reader base would translate to closer to 37,500 such readers! I hope that topics on engineers’ mind like paying down tech debt or adopting engineering practices are helpful to understand even at the product level.

The “other” bucket still accounts for a lot of people – closer to 75,000 folks, looking at the number of readers! Commonly mentioned titles included data scientist/analyst, security engineer, and recruitment. More sparse mentions included marketing, sales – and retired folks!

Types of companies

The publication covers Big Tech, startups and scaleups. Naturally, I was interested to see where readers work:

About 65% of readers indeed work at these types of companies. At the same time, there’s a healthy representation of people working at:

• More traditional, but tech-heavy companies

• Consultancies and developer agencies

• More traditional, non-tech heavy companies

• Bootstrapped companies (4.2%)

• Academia and research labs (1.8%)

• Public sector (1.5%)

• Nonprofits (1.1%)

• Other (~3%): self employed, founding a startup and in-between jobs were the most common mentions

Interests

I asked people for feedback on the publication, and gathered these observations: 

• Variation of topics makes the newsletter enjoyable. Many of you shared that you like it when topics are switched up. At the same time, the few times I did a multi-part series on a topic: I’ve received complaints from some of you how you’re ready to learn about other areas.

• The “practical research” pieces are a definite hit. Lots of professionals shared how they appreciate getting what feels like a closer to real-time feedback on what is happening in the market, and the industry.

• Disruptive technologies and companies: a common request. Many of you are asking to cover more about disruptions as we see them happen. Timely pieces like the ones on stacked diffs, modern developer productivity measurements, or companies like OpenAI are consistently popular, validating this ask.

4. Notable events

A few other things this happened in year three, that are worth a mention:

The Software Engineer’s Guidebook released

I started to write The Software Engineer’s Guidebook before starting the newsletter. In what is amusing irony: writing the newsletter delayed finishing the book – there’s only so much writing one can do, after all! Still, After four years of work, the book was finally released in print in November 2023, as an e-book in February 2024. The book became a #1 best seller across tech books on Amazon on launch. In the first 10 months, the book has seen more than 30,000 sales. Thank you to everyone who picked up a copy!

By the end of this year, I am aiming to release an audiobook version. Translations to 8 different languages are in progress (German, Japanese, Korean, Russian, Traditional Chinese, Simplified Chinese, Hungarian, Mongolian). The translations should be out between this fall and the spring of 2025 in respective countries.

I self published the book – as tech book publishers either passed on the title, or requested too many changes I did not agree with – and the self publishing process has been full of learnings. I’ll share more about this process, as self publishing, audiobook producing, and selling foreign rights has been full of learnings.

Leap Day

This year, February had 29 days. This small change caused a series of hiccups within software systems: from airlines to payment terminals. We shared a roundup of these. The next leap day will happen in 2028 – hopefully we’ll see more robust software running these critical systems by then.

The largest software-inflicted outage

In July of this year, we collectively witnessed the largest-ever software inflicted global outage where a routine Crowdstrike configuration update took airlines, banks and retailers offline across the globe.

We can now conclude that Crowdstrike did not do canarying or a staged rollout of this change, because they likely assumed that this change carries no risk. It’s always cheaper to learn from others’ mistakes: and this incident is a reminder that you should avoid doing “YOLO releases” when operating a critical line of business – even if that release is assumed to be safe to roll out.

5. Looking ahead

Some of the best-received articles come from guest writers: software engineers with deep experience in a topic. They then share their stories, learnings, mental models in a way that is helpful for more of us. Guest posts are always paid, and I closely work with guest authors to create a polished and enjoyable final piece. I’m always on the lookout for new guest writers:

Express interest in writing a guest post

One last thing… The Pragmatic Engineer Podcast. In the first two years of The Pragmatic Engineer, I ran several interviews with engineering leaders. A few examples:

• Platform teams and the platform team mindset with Ganesh Srinivasan

• Platform teams and developer productivity with Adam Rogal, Director of Developer Platform at DoorDash

• The full circle on developer productivity with Steve Yegge, formerly Head of Engineering at Sourcegraph 

The last 12 months, there have been no similar interviews – even though they are interesting and useful. At the same time, I kept talking with interesting people in the software industry, and frequently wished I could share these conversations wider. This is how the idea of the podcast was born.

The Pragmatic Engineer Podcast is an addition to the newsletter. Every second week, on Wednesday, a new episode will be out. 

This show covers software engineering and Big Tech and startups, from the inside. I do deepdives with experienced engineers and tech professionals who share their hard-earned lessons, interesting stories and advice they have on building software. 

After each episode, you’ll walk away with pragmatic approaches you can use to build stuff – whether you are a software engineer, or a manager of engineers. Expect a similarly polished experience as what the newsletter already delivers. Please subscribe using your favorite podcast player, or sign up the podcast’s YouTube channel or on Spotify or on Apple Podcasts. The episodes will also be shared in the newsletter, under the “Podcast” category.

I hope you’ll enjoy this addition!

For the next year, expect more deeply researched and practical deepdives both for technologies and interesting tech companies. If you’d like to suggest a topic or a company to research, you can do it here:

Suggest a topic

Thank you – and see you in the next issue!