How mass layoffs reshaped hiring, compensation, and the power dynamic between companies and engineers — told as a five-act tragedy that's still running.
🎭 Act I: The Binge (2020–2022)
To understand how we got here, you need to understand the party that came before.
In 2020, a global pandemic locked three billion people inside their homes and handed the technology industry the greatest growth accelerator in its history. Overnight, every company on Earth needed to be a software company. Restaurants needed apps. Schools needed platforms. Meetings needed to be Zoom calls, including the ones that could have been an email, which was all of them.
Tech companies responded the way tech companies respond to anything: by throwing money at it until the problem either went away or became someone else's problem.
Between 2020 and 2022, some of the largest technology companies on the planet doubled their headcount. Meta went from around 45,000 employees to over 87,000. Google parent Alphabet swelled past 190,000. Amazon — already enormous — crossed 1.5 million. Microsoft, Salesforce, Shopify, Stripe — the hiring spree was industry-wide, venture-capital-fueled, and spectacularly disconnected from any reality that would survive a return to normal interest rates.
And then there was the compensation.
During the Great Resignation of 2021–2022, over 47 million Americans voluntarily left their jobs. Engineers with two years of experience were fielding three competing offers before lunch. Sign-on bonuses hit six figures. Companies offered unlimited PTO, wellness stipends, remote-first policies, and in one case I personally witnessed, a "creativity sabbatical" that was literally four paid weeks to "explore your passions." The passion most people explored was interviewing for an even higher-paying job.
The employees had the upper hand. Companies knew it. Nobody questioned whether any of this was sustainable because if you questioned it, somebody else would offer your engineer $30,000 more and a ping-pong table.
Shaka, when the walls hadn't fallen yet.
📉 Act II: The Hangover (2023)
Then the money ran out.
Not literally — most of these companies were still printing cash. But the Federal Reserve raised interest rates, venture capital funding dropped off a cliff, and the magic math that had justified hiring two product managers for every engineer suddenly stopped mathing.
2023 was the year the industry sobered up, and it was not gentle.
An estimated 200,000 tech workers in the United States lost their jobs that year. Google laid off 12,000. Microsoft cut 10,000. Meta shed 10,000 on top of the 11,000 it had already cut in late 2022. Amazon, Salesforce, Dell, Cisco, IBM — the list reads like a who's who of companies that had spent the previous two years hiring like there was no tomorrow, and then discovered there was in fact a tomorrow and it had a budget.
Companies almost always gave some version of "we over-hired during the pandemic." This is corporate for "we made decisions based on vibes and venture capital and now we're fixing it by firing the people who had nothing to do with those decisions."
The cruelty was often breathtaking in its casualness. Engineers found out they were laid off when their badge stopped working. Some learned via a mass email sent at 4 AM. Others discovered it when Slack went silent and their calendar emptied — a phenomenon that became dark comedy material in every tech forum on the internet.
Average tech salaries declined for the first time in years. The Dice Tech Salary Report recorded the first drop in recent memory, with particular pain in management roles (down 14.8%) and cloud architecture (down 15.8%). For workers accustomed to annual raises that outpaced inflation by double digits, this was less a correction and more a rude awakening delivered by email from someone in HR they'd never met.
🧊 Act III: The Freeze (2024)
If 2023 was the layoff year, 2024 was the year of something arguably worse: nothing.
Around 153,000 more tech workers were laid off — Intel alone cut over 15,000, Tesla over 14,000, Cisco over 10,000 — but the bigger story was the silence that settled over the hiring market. Companies stopped firing. They also stopped hiring. The industry entered what analysts politely called a "no hire, no fire" equilibrium, which is a fancy way of saying "we're not going to lay you off, but we're also not going to replace your colleague who left, and yes, you're now doing both jobs."
Entry-level job postings fell off a cliff. According to Randstad, junior tech positions globally declined 29 percentage points since early 2024. SignalFire documented a 50% decline in new role starts for people with less than one year of experience at major tech firms. In Europe, junior hiring rates collapsed by 73%.
Let that number sit for a moment. Seventy-three percent. If you graduated with a computer science degree in 2024 and actually landed your first job, you should frame your offer letter.
The average time-to-hire stretched from 31 days to 44 days. Job seekers reported submitting 32 to 200 applications before receiving a single offer. The hiring process became a marathon of screening calls, take-home assignments, panel interviews, and then radio silence that stretched into geological time.
And then there were the ghost jobs.
👻 The Ghost Job Epidemic
This deserves its own section because it might be the most cynical development in the entire saga, and the bar for cynicism in this article is already remarkably high.
A ghost job is a listing posted by a company with no intention of hiring anyone. The position doesn't exist, was already filled, was never budgeted, or exists purely so the company can collect resumes "just in case." According to a 2025 Greenhouse study, between 18% and 22% of all online job postings in the United States are ghost jobs. Other studies put the number as high as 27%.
One in four to one in five job listings you see on LinkedIn right now is fake. Let that marinate.
A LiveCareer survey of 918 HR professionals found that 45% admit they "regularly" post ghost jobs. Another 48% say they do it "occasionally." That leaves exactly 2% of surveyed HR professionals who said they never post fake listings. Two percent. You have better odds of finding a unicorn in your backyard than finding a recruiter who has never posted a job they had no intention of filling.
Why do companies do this? The reasons range from "building a talent pipeline" (we want your resume on file in case we ever need you, but we won't tell you that) to "projecting growth" (we want investors and competitors to think we're expanding) to my personal favorite: "motivating existing employees" (we want you to feel replaceable so you work harder, and we achieve this by advertising your replacement before we've even decided to replace you).
Bureau of Labor Statistics data from mid-2025 showed 7.18 million job openings against 5.2 million actual hires. That gap — nearly two million listings that went nowhere — isn't entirely explained by ghost jobs, but they're a significant piece of the puzzle. When you subtract estimated ghost listings from the openings data, the real ratio of jobs to seekers drops from roughly 1:1 to something closer to 0.77:1.
There are fewer real jobs than there are people looking for them. The official numbers just don't show it.
States are beginning to fight back. Kentucky introduced legislation in January 2025 to ban ghost jobs outright. California passed a bill requiring employers to disclose whether a posting is for an actual vacancy. The FTC formed a Joint Labor Task Force with deceptive job advertising as a priority. Ontario, Canada enacted legislation scheduled for 2026 requiring timely candidate notification.
But for now, applying for tech jobs in 2025 involves a non-trivial probability that the job you're applying for is essentially a scarecrow. A mannequin in a store window wearing an "I'm Hiring" sign. A prop.
Welcome to the market.
⚡ Act IV: The AI Reshuffling (2025)
By 2025, the layoffs technically slowed. About 123,000 tech workers were cut — roughly 20% fewer than the year before. Progress, by the grim standards of the industry.
But the character of the layoffs changed. Companies stopped saying "we over-hired" and started saying "we're restructuring around AI." The framing shifted from correction to strategy. SAP invested €2 billion in AI while cutting 8,000 jobs. Dell eliminated 12,500 sales roles while increasing AI spending. Microsoft kept headcount frozen at 228,000 for an entire fiscal year while cutting 15,000 positions across multiple rounds — and then CEO Satya Nadella announced the company would resume hiring, but with a catch: every new hire would come with "a lot more leverage" thanks to AI tools.
Read that again. "A lot more leverage." The CEO of one of the world's largest employers publicly stated that the company's strategy is to hire fewer people and expect each one to produce more, because AI tools will multiply their output. This is the corporate version of saying "we're going to expect you to do three people's jobs, but we've given you a chatbot, so it's fine."
The salary picture reflected this new reality. According to Robert Half, tech salaries grew just 1.6% in 2025 — the lowest increase in at least 15 years. When inflation runs at 3%, that's a real-terms pay cut disguised as a raise. Silicon Valley saw a 7.3% salary decline. Software engineer salaries across the broader market dropped 9% to 15% depending on the segment. Top engineers from major companies reported accepting offers 30% lower than their previous compensation.
Equity packages shrank too. Carta's annual equity report showed that the average equity grant for new hires at startups declined by more than a third. Bonuses dropped from 39% of workers receiving them in 2023 to 31% in 2024. Training budgets were cut. Only 41% of tech professionals reported having access to learning or growth programs.
The message from employers was consistent and unmistakable: be grateful you have a job.
🔄 Act V: The New Normal (2026 and Beyond)
So here we are. Six years after the pandemic hiring binge, what actually changed?
The power shifted — hard. During the Great Resignation, engineers dictated terms. In 2026, only 28% of tech workers believe they're in a position to negotiate raises. Companies rolled back remote work, wellness perks, sign-on bonuses, and creativity sabbaticals. Jobs became transactions again, not lifestyles.
The junior pipeline is broken. Junior hiring cratered so hard it may take half a decade to recover. Companies are simultaneously complaining about talent shortages and refusing to train junior developers — the exact paradox I wrote about in "The Great Developer Famine." The companies that cut their junior programs in 2023 will be desperately trying to hire mid-level engineers in 2028 who don't exist because nobody gave them their first job.
Ghost jobs poisoned the trust. When a quarter of job listings are fake, the entire hiring ecosystem is corrupted. Job seekers burn out. Real postings get buried under noise. Companies that actually are hiring have to compete with the phantom listings for attention. Everybody loses except the companies gaming the system, and even they lose in the long run when their employer brand turns toxic.
AI became a negotiation weapon. Companies now use AI investment as justification for headcount reduction. "We're investing in AI" has become the 2025 version of "we need to do more with less," which was the 2023 version of "we're streamlining operations," which was the 2015 version of "we're laying people off but we'd like it to sound strategic." The technology changes. The euphemisms evolve. The outcome is the same: fewer humans, more work per human.
The compensation floor dropped. Software engineering is still well-paid relative to most professions. But the era of routine 20% raises, bidding wars between three competing offers, and equity packages that were basically lottery tickets is over. Mid-level engineers still earn well. Senior and staff roles still command premium salaries. But the ceiling came down, the floor came up, and the middle got squeezed.
🧭 What This Means for You
Here's where I stop narrating the disaster and start talking about what to do about it. None of this is revolutionary. Most of it is stuff you already know but haven't done because you're busy doing the work of two people since your colleague left and wasn't replaced.
🎯 Specialize or stagnate. The market is paying premiums for AI/ML engineers, cybersecurity specialists, cloud architects, and DevOps practitioners. Generalists are competing with every other generalist — and increasingly, with AI tools that can do generalist work passably well and don't ask for health insurance. The engineers commanding premium salaries in 2026 are the ones with depth in a specific, high-demand area, not the ones with a little bit of everything and a lot of opinions about architecture they've never implemented.
📦 Build your resume around outcomes, not tenure. Companies are hiring for impact now, not headcount. "I led the migration of 12 services to Kubernetes and reduced deployment time by 80%" gets you an interview. "5 years of experience with microservices" gets you added to a pile with four hundred other people who also have 5 years of experience with microservices. Every job you hold, document what you shipped, what you improved, and what you measured. Your next interviewer won't care where you worked. They'll care what happened because you were there.
🔍 Verify before you apply. Check when the listing was posted. Look for the "verified" badge on LinkedIn and Greenhouse. Check the company's Glassdoor reviews for recent hiring activity. If the listing has been up for 90+ days and the company announced layoffs last quarter, congratulations — you've found a ghost. Close the tab. Go for a walk. Your evening is worth more than a confirmation email from an ATS that will never send a follow-up.
💰 Negotiate from data, not feelings. Use Levels.fyi, Glassdoor, Blind, and Comprehensive.io to know the real ranges before you walk into a salary conversation. The company has done its homework on what they can pay you. Do yours. If they lowball you, it's not personal — it's the market testing whether you'll accept less because you're scared. Don't be scared. Be informed. There's a difference.
🛡️ Don't build your career on one company's stability. The engineers who survived 2023–2025 with their careers intact were the ones who had maintained their network, kept their skills current, and didn't confuse their employer's need for them today with any promise about tomorrow. Keep your LinkedIn active. Attend meetups. Do side projects. Write about what you know — start a blog, even. Not because you're planning to leave, but because the decision to leave might not be yours, and when that email from HR arrives at 4 AM, you'll want to have options that don't start with "update resume from scratch."
📚 Invest in yourself because your company won't. Remember that stat about training budgets being cut? Less than half of tech workers have access to any learning resources through their employer now. The other half are either learning on their own time or slowly becoming obsolete while their company saves $200 per quarter on a Pluralsight subscription. Be in the first group. The second group finds out they're in the second group when it's too late to switch.
📊 The Numbers That Tell the Story
For those who want the data without the commentary:
| Metric | Figure |
|---|---|
| Tech workers laid off, 2022 | ~93,000 |
| Tech workers laid off, 2023 | ~200,000 |
| Tech workers laid off, 2024 | ~153,000 |
| Tech workers laid off, 2025 | ~123,000 |
| Tech workers laid off, 2026 (through March) | ~53,000 |
| Total since pandemic correction began | ~622,000 |
| Average tech salary growth, 2024 | 1.2% |
| Average tech salary growth, 2025 | 1.6% |
| Silicon Valley salary decline (YoY) | -7.3% |
| Software engineer salary decline range | -9% to -15% |
| Entry-level posting decline since Jan 2024 | -29% |
| Junior hiring rate decline (Europe) | -73% |
| Ghost job percentage of listings | 18–27% |
| Average time-to-hire | 44 days (was 31) |
| Applications per offer | 32–200 |
| Workers who feel they can negotiate raises | 28% |
🪞 The Part Nobody Wants to Hear
The tech industry spent 2020–2022 pretending that infinite growth was possible, 2023 pretending the correction was a one-time event, 2024 pretending that "right-sizing" was about efficiency rather than broken planning, and 2025 pretending that AI makes it all okay.
None of those stories were true.
What happened was simpler and uglier: the industry hired based on optimistic projections, funded by cheap money, and when the money got expensive, it fired the people it had just hired. Then it spent two years cautiously re-hiring fewer people at lower salaries while posting fake job listings to keep the illusion of demand alive. Then it discovered AI and decided that was a good reason to hire even fewer people. And somewhere in all of that, an entire generation of junior developers got locked out of the profession because nobody wanted to pay for training when they could demand five years of experience for a first job.
Power shifted from employees to employers, and employers used that power exactly how you'd expect: reduce compensation, eliminate remote work, cut benefits, increase workload — while posting record profits.
It's not a conspiracy. It's just what happens when an industry that prided itself on disruption gets disrupted by the oldest economic pattern in the book: boom, bust, and the people who made the decisions don't bear the consequences.
If you're in this market — and if you're reading Codyssey, you probably are — the best thing you can do is be honest about the situation, strategic about your skills, and cynical enough about corporate messaging to read the job listing twice before you spend your evening applying for a ghost.
The market will recover. It always does. But it won't recover to 2021 levels, because 2021 levels were never real. What comes next will be different. More competitive, more AI-augmented, more demanding. The engineers who will thrive in it are the ones who stopped waiting for the party to come back and started building something durable instead.
📚 Sources
- Crunchbase News — Tech Layoffs Tracker (2022–2025 data). crunchbase.com/startups/tech-layoffs
- Layoffs.fyi — Live tracking of tech layoffs since 2020. layoffs.fyi
- TrueUp — Layoffs Tracker, 2025–2026 figures. trueup.io/layoffs
- TechCrunch — Comprehensive list of 2024–2025 tech layoffs. techcrunch.com
- Dice 2025 Tech Salary Report — Salary trends, regional data, role-specific compensation. dice.com/tech-salary-report
- Robert Half 2026 Salary Report — Tech salary projections, role demand. roberthalf.com
- IEEE-USA InSight — 2026 tech salary trends outlook. insight.ieeeusa.org
- Greenhouse — Ghost job research (2024–2025), 18–22% of listings are fake. greenhouse.com
- Clarify Capital — 1 in 3 employers admit to posting fake listings (2025 study).
- LiveCareer — 93% of HR professionals engage in ghost job posting (2025 survey).
- ResumeUp.AI — 27.4% of LinkedIn listings are likely ghost jobs.
- Bureau of Labor Statistics (BLS) — JOLTS data, unemployment rates, job openings. bls.gov
- Randstad — Entry-level position decline data (29% drop since Jan 2024).
- SignalFire — 50% decline in entry-level hires at major tech firms.
- Ravio — 73% decrease in entry-level hiring rates in Europe, 2025 Tech Job Market Report. ravio.com
- Carta Annual Equity Report — Equity grants down by a third at startups.
- Salesforce Ben — Year-end layoff analysis, 2025. salesforceben.com
- InformationWeek — Major tech layoffs tracker, updated December 2025. informationweek.com
- The Interview Guys — 2025 Job Market Year-End Review. theinterviewguys.com
- Congressional Research Service — Ghost jobs report, April 2025. congress.gov
- Microsoft/Satya Nadella — "More leverage" hiring strategy, BG2 podcast. techbuzz.ai
- WeAreDevelopers — Software engineer salary decline analysis (9–15%). wearedevelopers.com
Written for Codyssey by someone who has been on both sides of the layoff conversation and found neither side particularly enjoyable. The data is real. The sarcasm is a coping mechanism. The ghost jobs are, unfortunately, not haunted — they're just empty.