SW Tester Job Market Review — From the Human Side

What forum threads tell you that analyst reports never will — and how to actually listen.

"We eliminated the role."

That's how someone on Reddit answered when a 30-year-old career-changer announced she was hoping to land her first entry-level QA job in 2025. Three words, no softening. No "but here's how to make it work."

A few comments down, another voice: "QA is dying in the US. It's probably impossible to join now, especially without a CS degree and coding/automation."

And another, replying to a wife asking what her husband should do after eleven years in manual testing and a layoff: "I'd not recommend going into testing whatsoever."

If you only read those threads, you'd close your laptop and go retrain as a plumber.

Before I show you the other side — let me tell you what this piece actually is, because it's easy to misread. This isn't an article about statistics. I'll write one of those later: the BLS forecasts, the World Quality Report numbers, the analyst slides that project the QA testing market growing to a hundred and nine billion dollars by some year that always feels conveniently far away. That piece is coming. This one is different.

This one is about forums. About what testers actually say to each other, in their own words, in the middle of the night, when no recruiter and no marketing department and no LinkedIn audience is in the room. The Reddit threads where someone is six months into a job hunt and quietly tired of pretending it's fine. The Blind comments where someone is venting about a manager who just dissolved the QA team in a single Slack message. The replies under "is QA still a viable career" posts that range from compassionate to brutal to outright defeated, sometimes inside the same conversation.

These voices are subjective. They are sometimes disheartened, sometimes panicked, sometimes too quick to generalize from one bad year at one bad company, sometimes — let's be honest — too blind to a wider picture they cannot see from where they're standing. They are biased by who chooses to post and who stays silent, and the loudest emotion always wins the algorithm. In the strictest analyst sense, they are terrible data.

But here's the thing the numbers miss: the human. The person actually making the decision about whether to retrain, whether to keep applying, whether to give up on a career they spent twelve years building, whether to tell their kid that yes, software testing is still a good field to study. The numbers tell you what the market is doing. The forums tell you what it feels like to be standing inside that market, and what people are deciding to do about it. Both matter. The second one gets ignored almost completely.

So that's what this piece is gathering. Not predictions, not certainties. Voices.

But I kept reading. And on the very same forums, on the very same days, you find this:

"AI not being able to replace QA roles is an indication that the whole AI thing is a big bubble where everyone's claim about saving money is just a pipe dream and it's going to come crashing down."

"Manual testing isn't dead. Most companies now expect at least some automation skills — that's different from 'dead.'"

"SDET roles are growing rapidly. SDETs who are strong on the SWE side are very much in demand."

So which one is true?

Well — that's the wrong question. And falling for it is exactly what gets testers either panicked into a career change they don't need, or comfortable in a job they're about to lose. Both costly. Both avoidable.

Let me show you what I mean.

The pessimists are not lying. They're describing one room.

When someone on Blind writes "my company just replaced half the QA team with an automated testing suite and the mood in the office is dismal," I believe them. That happened, right now, in their building.

When another tester writes "my previous company laid off all manual QA in early 2023, and most have not found new work — the ones who have either pivoted to development or took significant downlevels," I believe that too. Also real, also painful.

When someone says "there are no more entry-level or mid-level manual QA positions in our IT org — devs are being forced to write automated tests, and it's going SO well… not" — I have spent enough time in enterprise QA to know exactly what room they're sitting in.

The thing is, none of these people are wrong. They're describing the company they work for, the team they were in, the recruiters who ghosted them. That's real data. It just isn't the whole map.

Here's what I noticed after going through a few hundred of these threads: the pessimistic posts almost all come from people who got hit. Layoffs. Six months of job hunting. Manual QA teams dissolved in a single Slack message from leadership. When you have just lived through something, the natural human move is to generalize from it. "My company did this, therefore the industry is doing this." It's not dishonesty. It's pattern recognition without enough samples.

And the loudest voices on a forum, by definition, are the ones with the strongest current emotion. The tester who's quietly employed and learning Playwright doesn't post. The tester who just got laid off after 11 years posts a lot. So the feed skews dark.

The optimists aren't lying either. But they're describing a different room.

The thread where someone says "automation testing and QA engineering still have demand, but roles are definitely shifting" — that's also true. The post where a senior SDET at a large fintech writes "we're hiring three more, can't find good candidates," also true. The Blind comment that says "QA at my company evolved into SDET, and the people who picked up coding are doing fine, the ones who didn't are gone" — true.

So you have two simultaneous realities. Manual QA roles are being eliminated. SDET and automation roles are growing. That's not a contradiction. That's a single sentence that the discourse refuses to keep together because nuance doesn't get upvotes.

There's also a third kind of comment, and these are the ones I save:

"It's crazy because QA is extremely important, but no one's hiring for it anymore. Companies seem to think they can get away without manual testing — and I'd credit a large portion of the drop in software quality to that."

That tester isn't predicting the future. They're describing a slow-motion mistake they're watching their industry make. Companies are cutting QA, software quality is dropping, and the two are related. That's not pessimism. That's diagnosis. And it's worth more than any analyst slide deck.

What the analyst reports get wrong

Right now you can find a report telling you anything you want.

"The QA testing market is projected to grow from $76 billion to $109 billion by 2025." Great.

"76.8% of testing teams have adopted AI in their workflow." Sure.

"60–70% of routine testing tasks will be automated by 2030." Okay.

These aren't false. They're just not what a tester needs to make a decision about their career. Because none of them tell you whether your role at your company in your country is safe in the next 18 months. The aggregate is fine. The aggregate is always fine. The aggregate doesn't pay your mortgage.

The Reddit threads, by contrast, are messy, anecdotal, contradictory, emotional — and far more useful, if you read them right.

How to actually listen

I'll be honest: scrolling QA forums looking for signal in the noise is not the world's most pleasant evening. But there's a way to do it that works.

Stop reading for conclusions. Read for observations.

When someone says "AI will replace all QA jobs in 5 years" — that's a conclusion. Ignore it. They don't know. Nobody knows.

When someone says "my company laid off the manual QA team and now developers are writing the tests, and the quality is visibly dropping" — that's an observation. Save it. That's information. You don't need to agree with their interpretation. The observation has value on its own.

When someone says "QA is dead, don't get into it" — conclusion. Skip.

When someone says "the last three interview cycles I went through, every QA role required Python and at least one CI/CD pipeline tool" — observation. Useful. That tells you what hiring managers are actually putting on job specs right now, which is a leading indicator of what skill gaps will hurt in 12 months.

When someone says "my SDET friends are getting offers within weeks, my pure manual-QA friends are at six months and counting" — observation. Massively useful.

The pattern, once you start filtering this way, is almost embarrassingly clear:

• The role label "manual QA tester" is contracting — especially in the US, especially at companies under cost pressure, especially for anyone with less than five years of experience.
• The work hasn't gone anywhere. Testing still happens. Bugs still ship. Quality is, by every honest account, getting worse.
• The work is being redistributed — to SDETs, to developers, to offshore teams, to AI-assisted tooling — and not always well.
• The people doing fine are the ones who can do the work under several different labels. Not because they're smarter. Because they're harder to put in a single box that someone can draw a line through on a layoff spreadsheet.

That's the whole picture. The forums told me. No statistic told me.

What this means for you, today

I'm not going to give you a five-point action plan. The whole point of this piece is that the specifics depend on your actual room — your company, your country, your level, your current skill stack — and a Medium article cannot know that.

But there are three things worth keeping in your head, the next time someone confidently tells you either that QA is dead or that QA is fine.

One. Fear is a reasonable response to genuine uncertainty. It's not weakness. It's how brains work when the map keeps changing. The thing to avoid is letting the fear push you into a decision before you've actually listened. Listen first. Decide second.

Two. Anyone with a strong opinion about what the AI shift means for testing jobs — including me — is, at best, extrapolating from a small sample. Hold every confident take, including your own, with a slight raised eyebrow. The honest version of every prediction in this space right now starts with "I think, based on what I've seen…"

Three. The single most reliable signal I've found is not on Twitter, not in analyst reports, not even in the loudest forum threads. It's in the dozens of quiet posts from people doing the work, describing what their team looks like this quarter compared to last. Those people are not selling anything. They have no horse in any race. They're just naming what they see. That's gold.

So, is QA dying?

It depends. It depends on what you mean by QA. It depends on what country you're hiring in. It depends on whether your team is at a profitable enterprise or a startup burning runway. It depends on whether you're 25 with bootcamp QA training or 45 with twelve years of test architecture experience.

What's not in doubt — and this is the only thing I'll be unhedged about — is that the people loudly declaring either funeral or all-clear are wrong. Both of them. They're describing rooms and calling them the world.

Read the threads. Filter for observations. Make your own map.

That's the work. There's no shortcut.