The software engineering industry in 2024: what changed, why, and what is next (Craft Conference)

The Pragmatic Engineer • • 44min → 9 min • #11
The software engineering industry in 2024: what changed, why, and what is next (Craft Conference)
Watch on YouTube

Summary

  • The software engineering industry in 2024 is undergoing a major shift that feels new but has strong parallels to the dot-com bust era of the early 2000s. Gergely Orosz, author of The Pragmatic Engineer newsletter and former engineer at Uber, Skype, and JP Morgan, delivered this keynote at Craft Conference 2024 to explain what changed, why it happened, and what comes next. The core argument: a 15-year period of near-zero interest rates and two massive technology revolutions (smartphones and cloud computing) created an era of abundant venture capital, rapid hiring, and engineering excess. That era ended abruptly around 2022–2023 when interest rates rose sharply, triggering layoffs, fewer IPOs, reduced VC funding, and a tougher job market—even at highly profitable companies like Meta, Google, Amazon, and Microsoft.

What changed in the tech industry

  • The job market flipped from extreme heat to contraction

    • In 2021, it was the hottest tech job market in history: candidates had multiple offers, companies struggled to hire even with remote work policies, and well-funded scale-ups lost candidates to competitors like Twitter.
    • By early 2022, layoffs began suddenly—first at unprofitable startups (Fast, Bolt, Klara, Flexport), then spreading to profitable giants. Meta laid off 11,000 people in November 2022 despite near-record profits; Amazon, Google, and Microsoft followed in 2023.
    • The layoffs.fyi tracker shows a massive spike in 2023, and the cuts continued into 2024 even as these companies posted record profits.

  • VC funding collapsed after a 2021 peak

    • Venture capital investment rose steadily from 2011 to 2021, with a huge spike in 2021. Since then, it has declined sharply.
    • As of Q2 2024, VC funding levels are back to where they were in 2018.

  • IPOs nearly disappeared

    • After a peak in 2021, tech IPOs dropped to zero in 2022 and only three in 2023. A few have returned in 2024, but the pipeline remains thin.
    • Fewer IPOs means less liquidity for investors and fewer exit opportunities for employees with equity.

  • Big Tech laid off staff despite record profits

    • Google is a striking example: founded in 1998, it had essentially no mass layoffs for 15 years (only a small one during the 2008 financial crisis related to the Motorola acquisition). In 2023 it cut 6% of staff, and more cuts followed in 2024—despite being repeatedly named one of the best places to work in tech.

Root cause #1: Interest rates

  • The Federal Reserve raised interest rates aggressively in 2022–2023, the largest hike in 28 years, to combat inflation that had reached ~8% in the US.

    • Higher rates mean borrowing is more expensive, consumers spend less, and demand cools—this is the standard mechanism central banks use to fight inflation.
    • The US went from 0.25% interest rates to 5% in less than a year.

  • For 15 years (2008–2022), interest rates were near zero globally—a historically unprecedented period.

    • Looking back to 1955, rates had only been at or below 1% for a few months at a time before. The 15-year stretch of near-zero rates was unique.
    • This was true not just in the US but in Canada, the UK, and the EU.

  • Startups are a “low-interest-rate phenomenon” (a phrase from Bloomberg analyst Matt Levine).

    • When rates are near zero, a dollar in 20 years is worth about the same as a dollar today. So investors are willing to fund startups that lose money for a decade in hopes of a big future payoff—because the alternative (keeping money in the bank) earns nothing.
    • When rates rise to 5%, investors can earn ~$5 million per year risk-free on $100 million. Suddenly, locking money into loss-making startups for 10+ years is much less attractive.
    • This is why VC funding dried up: it’s not that tech stopped being promising, but the opportunity cost of investing in startups changed dramatically.

  • Higher rates affect everything downstream

    • Less VC funding → fewer startups get funded, fewer IPOs.
    • Big Tech faces pressure to generate more profits (since shareholders can now get solid returns elsewhere).
    • The job market tightens as companies cut costs and slow hiring.

Root cause #2: The smartphone and cloud revolutions masked the end of an era

  • The timing of the smartphone and cloud revolutions coincided almost exactly with the start of the low-interest-rate period.

    • iPhone launched in 2007, Android in 2008. AWS launched in 2006, Google Cloud in 2008.
    • These revolutions created enormous new markets: Spotify, WhatsApp, Instagram, Uber, Snap—all exist because of smartphones. Netflix, Airbnb, Stripe, Twitch—all were enabled by cloud computing, which eliminated the need to build expensive data centers.

  • After 15 years, the smartphone and cloud advantage has diminished.

    • Building a mobile app or running on the cloud is now table stakes—everyone does it. It’s no longer a differentiating advantage.
    • The question now is whether AI will be the next transformative revolution, but it’s launching in a high-interest-rate environment, which changes the investment dynamics significantly.

The new reality for software engineers

  • It’s harder to get a job

    • Job postings on Hacker News, Indeed (US, UK, and Germany) peaked in 2021–2022 and are now back to 2019–2020 levels.
    • Supply Pike, a US-based scale-up, reported 192 internship applications per open headcount, 164 for software engineer roles, and 37 for senior engineer roles. Internship applications doubled year-over-year; software engineering applications tripled.
    • More applicants are coming from big tech layoffs (former Facebook, Google engineers).
    • Senior engineers are taking the first offer they get rather than shopping around.
    • Compensation negotiation power has shifted back to employers.

  • People are quitting less, which further reduces hiring

    • A survey of EU founders found that 50% saw no change in attrition, but the other 50% saw significantly fewer people quitting.
    • Less attrition means fewer backfill positions, which means fewer open roles overall.

  • Career growth is slowing

    • With less hiring and slower company growth, there’s less need for additional senior roles (tech leads, staff engineers).
    • Teams that used to grow from 10 to 15 to 20 people year over year are now staying flat, removing the natural promotion pipeline.

  • Shopify is preparing for fewer promotions by redesigning its career framework

    • Previously, Shopify used a simple level system (C1–C10) where advancement meant moving up a level.
    • Now they’ve introduced a “Mastery score” (0–50) at each level. Engineers can advance their Mastery score every 6 months—earning pay raises and recognition—without changing their level.
    • Changing scope (e.g., from team lead to discipline lead) is now treated as a different job, and your Mastery score resets.
    • This is a smart way to keep engineers engaged and rewarded even when promotions are rare. Gergely expects more companies to adopt similar approaches.

The new reality for software engineering practices

  • “Boring technology” is making a comeback

    • With less money and more pressure on efficiency, companies are more likely to choose proven, boring technologies (Java, PHP, jQuery) over exciting new ones (Rust, Next.js, Kubernetes).
    • Engineers may still want to use new technologies, but business leadership will push harder for boring, proven choices.
    • During the zero-interest-rate era, companies like Monzo famously adopted microservices as a hiring strategy—their CEO’s internal reasoning was: “Step 1: Go. Step 2: Microservices. Step 3: Who knows. Step 4: Profit.” Monzo ended up with thousands of microservices for ~500 engineers.

  • Monoliths are trending again

    • With less hiring, the organizational complexity that microservices solve is less of a problem.
    • Shopify has long used a “modular monolith” approach. More companies are starting with monoliths and planning to stay with them.

  • Full stack is becoming the default

    • A traditional small team building iOS, Android, web, and backend might have 5–6 specialized engineers. A full-stack approach using React Native or Flutter can achieve similar output with 3 engineers.
    • TypeScript is enabling this: Bluesky (a Twitter competitor) uses TypeScript across backend, frontend, and mobile (via React Native and Expo), so all 12 of their engineers can modify any part of the stack.
    • Linear’s CEO (a former Uber colleague of Gergely’s) said using TypeScript across the entire stack has been a major advantage.

  • Developer responsibilities are “shifting left”

    • Developers are taking on more work that used to belong to specialized roles: QA, operations, security, and even some project management.
    • Specialized roles (security engineers, SREs) will still exist but in smaller numbers per team.
    • Gergely does not believe AI will replace software engineers, but engineers will be expected to do more with the help of AI tools.

  • Less reinventing the wheel

    • During the boom, large companies routinely rebuilt internal versions of tools like React Native. This was expensive and often quietly retired.
    • In the new environment, companies are more likely to buy existing software or adopt open-source solutions rather than building custom platforms.

Haven’t we seen this before? Parallels to the dot-com bust

  • The dot-com bust (2000–2001) had striking similarities

    • There was a massive investment splurge from 1998–2000 into companies like Webvan (which promised 5-minute grocery delivery and went bankrupt). Instacart now offers 30–60 minute delivery and is successful.
    • After the bust, it was extremely hard to get jobs. Nitya Henry, now Director of Engineering at Spotify, graduated in 2001 and described competing with experienced engineers for any role. She accepted part-time and unpaid internships just to build her resume, pivoted to consulting and technical PM roles for 5–6 years, and eventually returned to software engineering.

  • Kent Beck (Agile Manifesto founder, author of Tidy First) sees the same patterns returning

    • Feedback loops are getting longer: teams don’t want to know about mistakes; they hand off and move on.
    • Fixed specifications and upfront design are back. Iterating constantly is happening less.
    • More documentation (as a way to show “we did our job, don’t bother us”).
    • More handoffs between developers and QA.
    • Gergely’s theory: in a stressful environment where people fear layoffs, there’s less incentive to take risks or fix problems outside your immediate responsibility.

  • Key differences between 2000 and 2024

    • In 2000, the internet was the emerging revolution. Today, it’s AI/LLMs—with 25% of all VC investment going into AI by mid-2023.
    • In 2000, the tech industry was much smaller. The seven largest publicly traded companies in 2003 included only three tech firms (Google, Microsoft, Apple). In 2024, five of the seven largest are tech companies, and they are vastly larger in absolute terms.
    • The tech industry is now deeply embedded in the global economy, so the impact of this downturn is much broader.

Takeaways and advice

  • The “doers” are safer than the “planners”

    • Engineering managers are being cut more than individual contributors as companies flatten teams. Many managers are moving back into tech lead or hands-on roles.
    • The push is to do more with less, use boring technology, and go full stack.

  • Getting a job requires more effort

    • Your network matters more than ever. Referrals are critical.
    • Apply early—some companies only review the first ~50 applications.
    • Take-home assignments and unpaid interview work are becoming the norm.
    • It’s becoming a numbers game: apply to many places.

  • Become product-minded

    • Understand your company’s business model: how does it make money? Can it become profitable?
    • Build relationships with product managers. Talk to customer support and research teams.
    • Engineers who can help the company make more money or spend less are the most valuable. Gergely wrote about this in his article “The Product-Minded Software Engineer.”

  • Aim for career security, not job security

    • A commenter named Alen May, who survived four rounds of layoffs in 2001–2002 (her organization went from 180 to 35 people), shared that she learned there’s no such thing as job security—neither as a manager nor as an employee.
    • What you can control is career security: keep learning, work on challenging projects, and work with great people. After being laid off, she found new positions again.

  • AI/LLMs are tools, not replacements

    • Gergely prefers the term “large language models” over “artificial intelligence” because it’s more accurate.
    • These tools make developers more productive—he uses Copilot and ChatGPT for brainstorming, pattern recognition, and picking up unfamiliar technologies.
    • His advice: understand how they work (he recommends Steven Wolf’s article “How Does Chat GPT Work”), try multiple tools, improve your own workflows, and avoid the hype. Most hype comes from people who don’t actually write code with these tools.

  • This has happened before, and the industry recovered

    • For engineers who weren’t working professionally in 2000, this feels unprecedented. But the tech industry has been through this cycle before.
    • What’s happening now is a return to how things were in 2018–2019 or even the early 2000s—and the tech world was still a great, interesting industry to work in then.
    • The industry remains unpredictable, but that unpredictability is part of what makes it challenging and fun.
Back to The Pragmatic Engineer