Coding Bootcamps Speedran Higher Education's Death Spiral

I’ve been building reskilling programs for over a decade. When I switched from software architecture to education, I spent years playing catch-up with learning theory, direct observation, and experimentation. In other words, I took an engineering approach to adult education.

One thing that always bothered me was that my business shouldn’t have existed at all. If higher education were affordable, accessible, and of high quality, there would be no need for non-accredited education companies like mine.

I remember when I pulled all of my retirement savings out and invested in my own startup. As a technical leader and hiring manager, the quality of graduates wasn’t getting it done, and damnit, I was going to do something about it. I had three goals for my new business:

1. I wanted to provide a genuine, quality learning experience for people who not only wanted to learn but were ready to approach skill acquisition like craftspeople.
2. I wanted to provide an affordable experience with a relentless focus on quality learning experiences and outcomes.
3. I wanted to show higher education that there was an opportunity to mix vocational, job-ready skills with the other skills that make someone a well-rounded individual.

The result? On the first two goals, I exceeded my expectations. We achieved a near 93% job placement rate, absolutely crushing the outcomes of both accredited and non-accredited alternatives at $10,000 per head. But the third goal I failed miserably at. Although we had a few higher education partnerships, for the most part, accredited universities wanted nothing to do with anything outside of their worldview.

When we were told “no”, it was usually one of these reasons:

• Faculty Resistance: They wanted to do things the way they always have. The introduction of a new program, new ways of operating, and new curricula was interpreted as a failure on their part. Ego trumped student outcomes.
• Rigidity of Higher Education: All learning, no matter what the topic or outcomes, must fit into the semester, trimester, or quarterly schedule. If quality learning doesn’t fit in that neat little box, it has to be stuffed with fluff or cut back enough to fit.
• Competition Concerns: Mostly competing with their own programs. “But if someone can get a good career at that price and velocity, it’ll cut into our revenue from traditional programs!” Yes, but only if your traditional program doesn’t have any additional value, but ok.

The bootcamp space erupted, and by 2018, there were millions of dollars in investments and coding schools in pretty much every major city. Most of these schools were backed by venture capital and run by business and marketing professionals, not technologists. In 2018, I left my company after it had been acquired and started raising the alarm.

Over the last few years, I watched the bootcamp industry move from innovative to completely hollowed out. They made the same mistakes that higher education is making, just on a faster scale.

The collapse is less about market saturation and economic cycles and more about suicide by a thousand compromises. Universities are following the same playbook.

How Things Fell Apart

Early bootcamp graduates were outliers. These were self-selected, highly motivated people who would have figured it out anyway. The bootcamp gave them structure, a forcing function, and lots of professional-grade feedback. Most of my students at the time either had no degree, because they hated all the performative nonsense in traditional schooling, or had a degree that wasn’t aligned with their career goals. For the latter, traditional education hadn’t worked out, and they certainly weren’t eager to pay for a second serving.

Regardless, high-quality, motivated, risk-taking students came in, and quality graduates came out. Employers hired them, and at least in my programs, they got excellent results. My hiring partners were pleased that my learners were not only technically competent but also that their diverse backgrounds outside of IT added value in communication, life experience, professionalism, and product design.

Across the industry, things were pretty good.

But then, the Growth Targets Kicked In

Investors are in the business of making money. There’s nothing wrong with that. However, there is a perverse incentive structure in scaling and growing education.

For starters, enrollment became the metric that mattered. You can’t hit huge enrollment targets with selective admissions, so the standards dropped. This is where the “Anyone can learn to code” grifters came in and turned an inspirational quote into a business model. Enrollment standards became performative, then were eliminated.

I'm sorry, but not everyone can learn to code. It’s as ridiculous a statement to make as “anyone can do brain surgery”. You need time, motivation, some aptitude for coding meta skills, and a genuine interest in the topic.

If you want to start a fight with me, dupe a bunch of people who aren’t prepared for technical rigor into taking on debt for a course that won’t deliver good outcomes.

With enrollment standards a thing of the past, the students who showed up weren’t ready. But you can’t fail people when your marketing and revenue model depends on completion rates. Yes, completion rates should be tracked, but they are less important than the outcomes of skill acquisition and employment.

So, what happens when you let people into an accelerated training program who aren’t ready for it? First, they start taking up all of the instructor’s extra time. Instead of spending effort pushing the high performers further and shoring up the middle of the pack's weaknesses, instructors are stuck doing remedial tutoring, watching their learners fall further behind through no fault of their own.

But, we can’t have dropouts because those completion rates and NPS scores matter! So, the rigor collapses. Projects become easier, and assessments become passable with minimal effort. The certificates being handed out became participation trophies.

Employers Figured It Out

Things took a turn when employers started figuring it out. The early bootcamp grads could ship code. The latter ones couldn’t. When I would get hiring managers to meet with my students, the pitch usually went like this:

Me: You hire from college programs?
Them: Yep
Me: How’s that going for you?
Them: It’s a pain in the ass! They are missing so many basic skills, like version control, unit testing, and how full-stack applications come together.
Me: So, if I provide you with a pipeline of junior talent who already know all that and can ramp up more quickly, would you be interested?
Them: Hell yes!
Me: Then there’s no harm in taking half a day to speedrun some interviews, right?

And that was that. They’d come in, almost all of them hired from me, and came back to hire more.

But around 2018, the narrative started to change. I saw more noise on social media from hiring managers blacklisting anyone who had learned at a bootcamp. Apparently, they were being flooded with applicants from people who had a bootcamp credential but no real skills. Sound familiar?

But the enrollment number still had to go up! So, it was time for creative financing! ISAs and loan programs delayed the reckoning. Students could defer the cost until they got jobs that increasingly weren’t coming. The debt masked the quality problem until it couldn’t anymore. When the bottom fell out, it fell hard. Lambda School became BloomTech, which became defunct. App Academy laid off staff. Smaller programs just quietly shut down.

We saw placement rates go from 70+% to 20% today.

The Root Causes

It wasn’t just one thing that tanked the industry, but here are some of the key causes of the collapse.

First was the belief that you could compress expertise into time. Ever wonder why every program, no matter which language or stack, was about 12 weeks long? That’s because it's what fits the cash flow model and sounds good in marketing, not because that’s how long it takes to develop competence in a stack of skills.

I can’t tell you how many times I’ve gone in to consult with an education provider about a program, only to have them tell me the budget and pacing before they know what the learning outcomes actually are. It is crazy to me!

Second, they confused information access with transformation. I’ve had the opportunity to review a lot of curriculum from failed programs, usually because I’m brought in to consult on the “fire sale” and determine whether there’s anything of value left. Some programs were charging 5-figures and running on YouTube videos and old docs. No, your Python lessons from 7 years ago do not have much market value today.

Third, the assessment model was almost theatrical. Automated tests would verify that the code produced the correct output. However, few programs were asking learners, “Why did you make that choice?” or “What are the tradeoffs of this approach?” You know, engineering skills.

Finally, they failed to respond to changes in the market and demand. Record the lectures once, scale infinitely, and never update your programs based on what will help graduates actually succeed in the market. No feedback loops, minimal accountability to outcomes. Just completion metrics, NPS, and revenue growth. This was particularly relevant when the market corrections began. With more competition for entry-level jobs, the bar needed to go up. Instead, most camps cut staff and increased marketing. The last gasp was many reaching for unproven AI tools to generate content and replace instruction. Here’s a hint: it doesn’t work.

The Higher Education Parallel

Universities have been making the same mistakes; it’s just taking them longer to reach the same outcomes because society at large still believes in the intrinsic value of a degree, though that looks to be changing. Here’s what I’m seeing in the higher education space:

• Enrollment targets are driving admissions decisions.
• Grade inflation is making degrees meaningless as signals.
• Student debt is masking the disconnect between cost and outcomes.

Colleges are optimized for volume. Bootcamps optimized for volume. Both are destroying the thing that made the degree valuable: a credible signal that this person can do the work.

A Return to Quality

I build programs that don’t fail by doing the opposite of everything above. This means rigorous intake: prework, behavioral interviews, aptitude assessments. Not everyone is ready right now, and pretending otherwise wastes their time and money. More importantly, it protects the signal for everyone who completes. Every graduate who can’t perform destroys the employer's trust in the next cohort.

Next, the timeline must fit the outcome, not an arbitrary constraint. If it takes six months to develop real competence, then the program is six months. The market doesn’t care how fast you trained someone if they can’t do the work.

My learning arc is show→do→reflect→assess. Deliberate practice with metacognition built in, it’s so much more than content delivery. Information is only the first step. As my father once said, “There’s a reason why intelligence and wisdom are two different stats in Dungeons & Dragons.”

In technical fields, intelligence is wielding the information, and wisdom is knowing where it applies. Engineers, the ones you want to hire, have both.

AI Makes this More Urgent

Every assessment that LLMs can game is now worthless. Every learning experience built on busywork and performative exercises is now obsolete. You cannot hide behind multiple-choice tests or coding challenges that can be solved by pasting the prompt into Claude Code.

In a way, the destruction that AI is wreaking on higher education is a relief. Because now, the only thing that matters is whether or not someone can actually think. In IT and engineering disciplines, this means they understand the reasoning, can explain trade-offs, and defend their decisions. These are the things that universities used to test for, and most bootcamps skipped in their quest for “number must go up”.

The institutions that survive the next decade will be the ones that optimize for genuine skill development and rigorous assessment of understanding. Everyone else is selling credentials that the market has already stopped valuing.

The Reckoning is Coming

Bootcamps showed us what happens when you optimize for growth over outcomes. Universities are bigger, slower, and more entrenched. The degrees carry more prestige. The debt loads are higher. But the fundamental dynamics are identical.

The market is already correcting. Employers are building their own training programs and hiring based on demonstrable skills rather than credentials. Degree requirements are disappearing from job postings. The same thing that happened to bootcamp certificates is happening to bachelor’s degrees, just on a longer timeline.

As an industry, higher education can either learn from what killed the bootcamps or follow them off the cliff. The choice is whether we want to speed up the collapse or actually fix the problem.

The bootcamps taught us a lesson. AI is pouring gasoline on the fire. Rigorous intake, deep learning with real assessment, and constant improvement based on results are the way forward.

Whether or not higher ed is willing to embrace it remains to be seen.