Hacker News new | comments | show | ask | jobs | submit login
Why Are Coding Bootcamps Going Out of Business? (hackeducation.com)
133 points by imartin2k 8 months ago | hide | past | web | favorite | 128 comments

Because those who can learn under those conditions can easily learn the material for free and many have before attending. What they use a bootcamp for is a bridge into a career. People are finding other less costly ways into the job market.

The bootcamps of the future will focus on either the experienced developers wanting to learn a new language/skill quickly. The structure will be like an extended conference where everyone attends presentations and gets excited for a few days and then everyone goes home. This will be followed up with an online course teaching the basics that will allow everyone to work at there own pace. A few tests and assignments will assign grades. A hackton will be planned where everyone will build a product using there new skills. Teams will be assigned based on grades to make things fair.

I think people underestimate how difficult it is to teach yourself, especially when there are so many different ways to learn and a million different paths to take. One of the most important parts of bootcamps, in my mind, is that all of the curriculum is there, in the right order, and on a set schedule. When I taught myself to program I probably spent 80% of my time finding the right resources and understanding what I should be learning and 20% of my time learning. I think it takes a pretty special person to self-teach when you don't know what you don't know.

>I think people underestimate how difficult it is to teach yourself,

I wanted to give some counterpoint to that. The point is not to argue with you because you clearly felt that it was "difficult" so by extension, it makes sense to project the learning obstacles out the world so we know they are there. You felt there were hurdles and that's all in the past so we can't argue with what you felt.

(Hopefully, the above was said with enough empathy to not sound critical because one has to be open-minded about the following observation...)

There are others[1] who think that the discipline for self-teaching is the baseline skill for average computer programmers and that attending "coding bootcamps" is a negative signal. I believe the thinking is... if one requires coding bootcamps to learn Javascript/Ruby/Mongo, it doesn't instill confidence in the hiring manager that the person will be able to learn new topics on-the-job that will be more difficult, with an undefined curriculum, under tighter deadlines.

If people still want to say "fuck off" to the types of people that look down on coding bootcamps, that's fine as long you're aware that the hiring prejudices are out there and you have consciously chosen to ignore those companies.

[1] examples:




My point was not that the material itself was difficult to learn, but that knowing which direction to go in, or knowing what material to learn, was the more difficult part.

>, but that knowing which direction to go in

But the coding bootcamp's curriculum and sequencing isn't necessarily superior. It could be as arbitrary or suboptimal as any low-rated 1-star book on Amazon.

A curious programmer would actually be skeptical that the bootcamp has the proven syllabus for learning.

Therefore, find a 5-star book on Javascript/Ruby on Amazon, and start with Ch 1. What evidence (not marketing materials) would the novice programmer see that makes the coding bootcamp's sequencing superior to that?

It's not that it's superior, it's that it exists. Now I could easily go back and burn through what I learned in 1/10 the time, but in the beginning I didn't know where to start or what I needed to learn, and spent an insane amount of time flailing and jumping around. Imagine not knowing Java from JavaScript.

You get caught up in all the arguments about which language/framework/tech of the day is the best because you don't know any better. For example, my initial Google search led me to learn Haskell because it's "the best language", using Vim because it's "the best editor." Looking back that's silly and absurd, and most people would agree that's not a wise way to start (inb4 'Haskell is actually...') but I didn't know that then. You're flying completely blind.

Plus you are ignoring the social aspects of learning.

Straight book-learning is not a fit for everyone.

I hire junior engineers. When I am looking at a resume, I am looking for something that indicates success possibilities.

Many times I see the only programming experience as the bootcamp.

A 4 year CS degree means that the candidate had to stick with the CS for 4 years: 12x longer than the bootcamp. (I am assuming summer internships as well)

This time allows concepts to sink in, be reenforced by later classes, implications of choices to all be seen.

Human beings require a certain amount of time. Its not just the 10000 hours (or whatever) of training. I feel that it is the duration of time that the learning takes place that is also important.

Bootcamps are too short for the deep, deep learning needed.

I haven't yet interviewed a bootcamper (only) candidate that can answer questions beyond the basic phone screen. My only guess is that they have had all this knowledge thrown at them, and their brain hasn't yet sort things out.

Why take a candidate like this if I can get a BSCS candidate? And if I can't get a CS grad? I rather trim the work. I am not in the business of giving a 4year CS education.

You're comparing apples to oranges - of course 4 yrs is better than 3 months. Consider the same person at some point in the future, and it would be bootcamp grad + 3 years experience vs CS grad.

You could make a case for both at that point, but I'd guess many would be inclined to opt for a bootcamp grad + experience, all things being equal.

But you're suggesting that the employer should be paying for the remaining 3.75 years of experience. I'm certainly not going to hire someone who's going to require that sort of investment in comparison to a fresh college grad. You may as well just grab seemingly clever people off of the street.

While a bootcamp graduate would require more investment than a college grad, it wouldn't be anywhere as high as 3.75 years, because college teaches you a lot of skills that aren't directly useful (Turing machines) or not in your domain (compilers or image processing) or not for several years. Training customised for your needs will be much more time-efficient.

Yup, that's one of the flaws of bootcamps. But that said, there's lots a bootcamp grad can do with relatively little training, and some of that are things that CS grads may struggle to do day one. A college grad, of course, will require some training too. Just a different level of it.

It's difficult to make an argument for bootcamps, though, when they won't teach you most of what you'd need to know either. The curriculum is typically focused around a single tech stack (many teach Mongo for unfathomable reasons), leaving students ill prepared for the workforce. Debugging? How to sort a lidt efficiently? What is a hash? All of these are arguably critical to understanding simple, common code; you'd be lucky to make it through a year as a junior dev without being fired if all the skills you have come from a bootcamp.

Of course, Galvanize et all just hire their own grads to inflate the post-graduation job placement stat. So maybe it's not necessary to know anything outside the course to get a job!

You have a point. Here's the problem the bootcamps will start to run into:

They're teaching as much as they can in twelve weeks, but lengthening the program increases the cost dramatically. Why Mongo? Because they're already using JS so they don't have to teach two languages (no time), and mongo is super easy to pick up. They'll learn other stuff on the job. Why aren't they teaching how to debug and what a hash is? Well we have to choose between that and understanding what a database is, so you take your pick.

They would love to have classes be six months or a year, but then they'd have to double their prices (eliminating a lot of the demand) and people would have to support themselves with no pay for twice as long (eliminating most of the rest of it). It's hard enough finding people to take 12 weeks off and pay $10k, and now we need to charge $20k for six months? No one can afford that, what we have now works well enough.

In order to do something longer and more thorough/in-depth you'd probably have to start with a fundamentally different business model.

>In order to do something longer and more thorough/in-depth you'd probably have to start with a fundamentally different business model.

Interesting point. Maybe something in between bootcamps and full-fledged 4-year CS or Software Engineering degrees? Say a 2-year course, with some of each of: tech stack topics, algorithms, data structures, and other CS-y topics, and software engineering topics. That would make for more well-rounded junior devs than the trend of just teaching the hot tech stack du jour plus git plus some DB (and all at very basic level, as many of them seem to do).

Some Indian software training institutes (such as Aptech and NIIT) already did have long courses (with corresponding high fees), but to the best of my knowledge, they may not have had much in the way of the higher-level topics in those courses, i.e. CS and s/w engg. ones - AFAIK they mainly had the hot language and DB of the season, like Java and Oracle some time ago, now probably something else like JS and Mongo, etc.

Ya, I think so. I'm a co-founder of https://LambdaSchool.com/computer-science and that's part of our goal

Longer programs would also compete with flexible programs from traditional institutions. For example, Mills College's MA in Interdisciplinary Computer Science [0]. The traditional institutions arguably have large advantages in accreditation, financial aid, quality control, social status, signaling, and branding.

[0] https://www.mills.edu/academics/graduate-programs/computer-s...

Very true

Classic internet.

The argument for bootcamps is easy! It costs a FRACTION of a college degree, takes a fraction of the time, and almost guarantees you a job.

"you'd be lucky to make it through a year as a junior dev without being fired if all the skills you have come from a bootcamp" LOL.

I know literally hundreds of people who have done just that, and it aint luck.

There is tons in a comp sci degree that you don't need to know right away to be an effective working programmer. And lots of bootcamps do teach "How to sort a lidt(sic) efficiently". It's a little hard to generalise because bootcamps vary in quality - like anything - but they give you a good structured introduction to lots of relevant material and the tools to keep learning as you go.

>The curriculum is typically focused around a single tech stack (many teach Mongo for unfathomable reasons)

Could be because many startups used to use Mongo. Not sure about current state of affairs, but somewhat recently, plenty of them did. One of the stated reasons was "schemaless", and other might have been scaling. I don't necessarily agree with either, because it can depend on a lot of factors, and scaling may not be the first thing that a startup should plan for. Just saying what I've heard/read.

>how to sort a list efficiently


There will always be some types that are naturally inclined towards development. I learned on my own, no college degree, in the late 90's, when there was no Stack Overflow or Github. I lived in a tiny town, so there wasn't a local Barnes and Noble or the like.

I would say that someone who needs a curriculum won't succeed as a developer. They may remain hired, an eternal junior programmer, and as long as they're surrounded by seniors they'll make commits on a daily basis. They'll become the "one year of experience ten times" as they get on in their career. Succeeding as a developer really requires a knack of learning to learn and independent problem solving.

It's okay to not be at the upper level; we always need assembly line programmers. However, all of the marketing on bootcamps lumps everyone together, and doesn't tell the true story: there may not be enough assembly line programming jobs to go around.


> I would say that someone who needs a curriculum won't succeed as a developer.

I disagree.

Lets try this shall we:

"I would say that someone who needs a curriculum won't succeed as a "

... doctor

... pilot

... nurse

... electrician

... architect

I am not willing to accept a self-taught brain surgeon.

Nor would I. There's a reason why we're talking about this on HACKER News.

Every field you mentioned is a licensed profession, in an industry surrounded by regulations and inspections. Every one of those fields involves the health or safety of other people. Programming is a pretty wide open field, without regulation and highly prescribed ways of performing the work. It feels extremely disingenuous to compare a brain surgeon to someone building Uber for Skateboards and deciding whether to use Redux or observables.

Programming flaws kill people. WannaCry shut 20% of britian's hospitals. ( http://www.telegraph.co.uk/news/2017/05/13/nhs-cyber-attack-... )

Seems like programming DOES involve the health and safety of others.

Are you really comparing brain surgery with "let's write another CRUD app"? Every field you mentioned is licensed for a reason. Your Angular app isn't going to kill anyone, and 99% of devs will never touch anything safety critical.

What era was it when you are learning?

Now we have excellent community-developed free and open resources like https://www.theodinproject.com .

Years ago. It probably has changed to some degree.

But for example, what you just linked - that's a great resource, but it still doesn't get you very far; you'd have to jump from there to somewhere else to learn what a bootcamp would teach.

I learned by picking up a book on development. Its curriculum was right there, laid out in the correct order and I could work on it at my own pace.

These books still exist. But today there are also guided sites like code academy that do a great job at providing elementary resources for learning to program (though, I generally find the pace to be a bit glacial).

There are so many things wrong with this that I don't know where to start.

First, to say there's no skills gap is insanity. I refuse to believe anyone in the industry actually believes that - even the worst bootcamps are placing folks, they just place folks that are not well prepared and shift the burden of training to the company. I'm constantly blown away that some of those people get jobs, and they somehow keep them, as the companies just start training from almost scratch. That really sucks, and a lot of companies are (rightly) wary of that, and a lot of companies throw bootcamp resumes straight in the trash can, but there are enough that won't that it doesn't matter.

Don't get me wrong, most of the skills gap is for capable engineers, and if you understand data structures, algorithms, architecture, memory management, etc. you'll get a job very quickly. We're talking within a week. But most bootcamp grads don't know any of that stuff, so the market is flooded with junior rails devs. (And even those folks still slowly get OK jobs).

There are audited, standardized reports from an organization called CIRR. Look at them - you'll see 90%+ placement rates, audited and from third parties. Granted there are a lot of lies on the home pages of bootcamp grad websites, but the perception HN has is much worse than reality at the good bootcamps.

I know the internal financials of quite a few bootcamps, and most are making a killing. It's really not hard to make money by charging $10,000/head so long as you keep demand. Iron Yard got out ahead of its skis and tried to open 16 locations at once. I have no idea what happened to Dev Bootcamp but the only way they could be losing money is if their expenses were absurd. They were bought for $80 million. I think the most likely scenario is the acquirer tried to change things up and killed everything.

>Don't get me wrong, most of the skills gap is for capable engineers, and if you understand data structures, algorithms, architecture, memory management, etc. you'll get a job very quickly. We're talking within a week.

Empirically, that's totally false. The people I knew who went to the same bootcamp (App Academy) as me and had the strongest skills took longer, not shorter, to get jobs. The average was around 9 months.

The opposite was true for my cohort. There seems to always be at least one or two of the top students that are dead set on making as much money as possible so they take the extra time to brush up skills and projects to land interviews with and be successful at the absolute top tier companies (Google/Facebook + Unicorns).

a/A also really pushes the idea that students should leverage offers against each other. I suspect that combined with the success of Haseeb's article is that top students are now more likely to push very hard to get top dollar for their worth instead of taking ~10k less and getting to work sooner.

How deep doesn't App Academy go into stuff like data structures, algorithms, memory management, architecture, scaling, etc.? I see mention of algorithms on the site, but not much else that would lead me to believe developers understand more fundamental CS concepts.

Not very; about 10 hours is spent on that with instructors, plus more time practicing it in the form of whiteboard questions. Which means there was a wide variance from most knowledgable to least.

Ya that'll do it

As deep as you can go in 13 weeks while also covering tooling, best practices, and example projects in 2 languages.

i.e., not very deep.

Is there a good way to learn those skills at a professional level outside of a university setting? I feel like I've taught myself a sufficient amount of front end, back end, ops (Docker, IaC etc), but I'm at a loss for how to test my ability with algorithms and data structures, memory management, etc. I've picked up some books, but I don't have a good way to measure my ability or demonstrate it to employers.

I'm a self-taught web developer, and I can tell you that the skills and knowledge gap between web development and "real" software engineering is huge, enough to where I seriously consider going back to school to pick them up, even though I already make $100k+, I've reached the top of my earning potential.

If I knew a program tailored to people like me that would give me a reasonably-accredited certificate and teach me the math and theory, I'd pay perhaps $20k for it. I'd go back to college, but I really don't want the pain of all that undergrad coursework.

Have you considered the Georgia Tech online masters in comp sci?

Do you need a degree to do that? Or can you test your way in?

The free way is of course with books (http://teachyourselfcs.com is a good guide), but if you're looking for a school you could do something like bradfield or something like https://lambdaschool.com/computer-science (disclosure - I'm a cofounder of Lambda School).

As far as demonstrating it to employers, that comes down to the interviewing process, which is in and of itself pretty broken, but they'll probably just ask you a lot of contrived questions about that stuff.

Create some open source projects in the field you are interested in! This is a great way to demonstrate that you know what you're doing, especially if others are using your project and collaborating with you. When I interview people I always look at their GitHub and dig into their work. Just make sure that your code has tests and has ok documentation.

Bonus: if one of them gets popular you'll have companies coming to you.

Learning algorithms for the sake of it is incredibly dull, for me. I find reading about nice algorithms interesting, but I don't remember anything later unless I'm solving a problem.

You're better off working through a problem set like Project Euler, or trying some TopCoder Single Round Matches (SRMs, there's a huge archive). At least that way you'll see some interesting puzzles and you'll get exposed to the majority of the 'standard' algorithms.

By the time you get 'good' you should be able to read a problem statement and immediately think "Ok, this is some kind of graph problem" or "List of sorted stuff? Probably need to use binary search". There are some good books like Skienna's the Algorithm Design Manual which encourages this kind of intuition.

> but I'm at a loss for how to test my ability with algorithms and data structures, memory management, etc

Try to work on projects involving encoding/decoding video streams and displaying them on the screen.

Also, 3D game engines.

> There are so many things wrong with this that I don't know where to start.

Exactly my thoughts. The whole premise if false. It conflates people coming fresh off a coding bootcamp with "competent programmers". It's not even funny.

"They were bought for $80 million."

AFAIK that's not public info. How'd you find out about this?

why do you seem to think that training is something the companies should avoid if at all possible?

It's not something they should avoid, it's just something many are not currently prepared to do.

People are realizing that even though coding pays well, it's not a job for everyone. We hired 4 people out of boot camps. All had a decent level of skill (of course they did, we choose the 4 we liked out of dozens available). However after a few weeks, two of them left, saying: "there is no f'en way I can do this 40 hours a week for the next few decades". And they went back to their old jobs and old pay.

> "there is no f'en way I can do this 40 hours a week for the next few decades"

Not a bootcamp grad, but I just hit my two year mark a couple weeks ago and want to blow my brains out.

Writing internal web applications is the most mind numbing thing ever. It's gotten to the point where I have so much trouble keeping focused at work.

The real issue is that I have no idea what field I'd pivot into.

Look, writing software is a trade. If you're not getting satisfaction from the act of creating something and doing a good job of it, even if that something is less "amazing" than you were hoping it would be, you probably need a different career. In any skilled trade the majority of the projects you do will be of the more mundane sort, but a skilled practitioner derives satisfaction from the doing and the mastery. You wouldn't work for years as a watchmaker, hunched over a bench 8 hours a day, if you didn't get satisfaction from making watches. Programming is more analogous to that than not, imo.

>Look, writing software is a trade. If you're not getting satisfaction from the act of creating something and doing a good job of it, even if that something is less "amazing" than you were hoping it would be, you probably need a different career.

Maybe. I've been doing this for about 12 years now and, if I were writing typical IT/web dev stuff, I'd move to a different field. Most software is just boring and unrewarding, the money wouldn't keep me if the interest fell away (though I have a family to consider at this point, but that's beside the point.)

First, with the caveat of obviously not knowing you or your job, it's my experience that it's rarely the case that an area as broad as "writing internal web applications" is actually inherently boring to a person. The parable of the stonecutters come to mind, are you cutting stone or are you building a cathedral? Find out what energises you. Think back, what were good days and what did you do? Try to do more of that, try to make the work you do more like that kind of work.

Of course, there are plenty of jobs that aren't remotely in the cathedral building business, and if that's the case, start networking. Get on Meetup and attend some meetings or hackathons or startup weekends or whatever. Your skills are generally in demand, so focus on getting inspiration for what might interest you, or meeting people that you think you'd enjoy working with, and let that guide your job search.

And don't let internal webapps scare you. Building a "boring" app for a team that does good things with it and appreciates your work can be very satisfying.

Have you considered a different company? I worked at a large company for a few years building internal tools and I wanted to kill myself. I couldn't imagine working there for the rest of my life. But I enjoyed writing code in my free time and knew it was what I wanted to do. I've been at a smaller company for ~5 years and I couldn't be happier.

If it's mind-numbing, can it be automated? Can you use higher-level tools to build them and reduce the repetitive parts?

If only there were the right high-level tool -- an android with an engineers face! It could attend the pointless meetings, and answer QA's trivial questions, the same over and over. Then I could have stayed on at FANGcorp.

You may enjoy a product manager, technical program manager, or business analyst role more. Not sure what city you're in, but there are plenty of courses at Galvanize! Good luck and please do not blow your brains out!

It's gotten to the point where I have so much trouble keeping focused at work.

Are you sure you're not expecting too much of yourself? Keeping focused is pretty difficult for programmers in general, we often need breaks to recharge. Famous Joel Spolsky essay related to this called Fire & Motion:


[But if you're not enjoying the work, that's different. Life's too short for that, find something that you enjoy.]

Have you tried seeing a doctor or therapist? I have the same feelings as you and I'm giving SSRIs another go. There's a difference between hating your job without being able to cope and hating your job while being able to cope. Have you considered that you're depressed? Are you able to enjoy things and be productive outside of work or do you feel the same way you do as when you're at work?

It can be a hard road to travel and there might not be a single good answer (only multiple possible answers with varying probabilities).

>There's a difference between hating your job without being able to cope and hating your job while being able to cope.

The difference being in the latter case you've given up all hope and just keep at it?

Fifty percent of people you hired who came with experience of different industries told you that working conditions (be that the work itself, or your particular environment) were so bad that they'd rather go back to what they were doing before.

Did you take a look at the working conditions?

That sounds to me like a resounding success. 50% of your hires from a 12 week program worked out? Do you realize what that would mean for the country if we could put people back to work that cheaply?

thats the graduation rate of some high schools

I think the best bootcamps do a pretty good job of having students answer that question during admissions or during the class itself. If you aren't getting some fulfillment during class, you probably won't during work either. For what it's worth, my company has hired half a dozen developers out of bootcamps (Fullstack and App Academy) and none of them have left or expressed interest in leaving the field.

I don't know. I could easily see myself being the type of person who would have a lot of fun (and get a lot of fulfillment) during a bootcamp, but be bored out of my mind shortly into my first real job. I get a lot more enjoyment out of learning new things than I do from pure programming. I've been lucky that through my career I've gotten to work on some amazing projects at top of the line companies, so there's a decent amount of learning happening day-to-day -- far more than I am sure happens at run of the mill in-house developer jobs. But even then, I found that I dislike jobs that involve programming all day long, so I gradually transitioned into other roles where I can still use my technical skills but don't spend most of my time writing code. I definitely didn't have that clarity of preferences until I'd actually been doing the real job for a little bit, because it's way different from class.

I suppose. For New Economy jobs, they are at least an option. I had a friend who was toiling away at a thankless PR job, communications major. Did coding bootcamp and she is much much happier working at a financial/web services startup

Any specific feedback? What is it that they didn't like?

Perhaps a rather cruel take on the issue, but I think the only thing worse than mercenaries, is in-experienced mercenaries.

I think the biggest factor causing these bootcamps to go out of business was stated in the article: there is no IT worker shortage. The fact that most of the people attending those bootcamps already held Bachelors degrees strongly implies that. If there was a real worker shortage, employers would be hiring people without degrees and have them do the best they can.

I think bootcamps arose as a result of an industry-wide PR move to give the illusion of a worker shortage, done in an effort to push more people into the tech industry and increase the labor pool significantly, thus pushing wage growth below inflation. With the labor market finally actually tightening, and with the PR move not working as well as was hoped, people are starting to see that the virtue signalling offered by bootcamps aren't as necessary as they thought.

> there is no IT worker shortage.

There is no junior worker shortage. A bootcamp grad is competing with other bootcamp (100+ boot camps) grads, and people who just self started and got to coding.

Finding mid-level and senior people is actually quite challenging.

"Finding mid-level and senior people is actually quite challenging. "

Strangely finding a job is also quite challenging for senior people....

I think there's a market for lemons problem at play. Mid and senior developers who are effective at what they do are incredibly valuable and valued and therefore don't generally have to look via the main queue of job applicants. They often get retained well at their current place and are poached/referred more often than the average or ineffective developer.

(This breaks down when a company suddenly closes or does a layoff so deep that it puts a lot of good people on the street.)

The flip side of the very effective using a different mechanism is that the people in the main pool have a harder go of it, as hiring managers are used to seeing a parade of people who interview poorly (and therefore interview A LOT) or who join and fizzle out (and therefore job hunt A LOT).

> Strangely finding a job is also quite challenging for senior people....

Are we talking about senior experience or X years in profession? I often see seniors (and mid, but less pronounced) who learned a job and did basically the same job for many years. When it's time to find a new job, they are barely beyond junior skill wise.

Senior people who kept learning through their career are incredibly valuable and rarely seem to ever hit the open market. They end up recruited/poached from existing jobs.

I think this maybe a network issue. For those of use without a network there is more work finding a position, passing HR screens and interviews.

There's lots of work for people with skills. Worker shortage under one perspective is a logical impossibility - it just means salaries are too low - but from the other side, there's plenty of things that could be automated that aren't because salaries are too high, hence a shortage.

This is all aside from the difficulty of grading devs to establish different salary levels, of course. I personally don't think my salary is too high ;-)

> but from the other side, there's plenty of things that could be automated that aren't because salaries are too high, hence a shortage.

That just means there's a market opportunity out there to build a tool that helps accomplish this. The majority of companies in the world could not afford to write and in-house version of Quicken, but they have no problem paying a few grand a year in licensing fees for it.

>"With the labor market finally actually tightening, and with the PR move not working as well as was hoped, people are starting to see that the virtue signalling offered by bootcamps aren't as necessary as they thought."

The US economy may be moving towards full employment but that doesn't in any way mean that the tech sector is moving towards full employment. Tech is but one sector of the overall labor market. Tech along with health care are two segments of the economy that continue to add jobs faster than any other segment. See:


> If there was a real worker shortage, employers would be hiring people without degrees and have them do the best they can.

Doesn't this kind of assume that even someone with very little skill can add value to a company as a software developer? It might be possible that below a certain skill level, they would make things worse instead of better.

Is there evidence that wage growth in the IT sector (or for programmers specifically) is below inflation?

I don't see that at all personally (as an employee nor as a hiring manager).

I agree with you, and keep trying to figure out where all the doom and gloom people are coming from working in IT. Maybe it's everyone in SF dealing with insane housing prices? Everywhere else, programmer salaries are some of the best in a given location and continuing to rise. Job openings are also on the rise.

A 250k Haskell programming job in SFBA is prima facie evidence of a "skills gap". The law of supply and demand says that's how you evaluate an inadequate number of available programmers. It's facile to suggest there's an adequate labor supply because employers can acquire developers simply by paying 5x the national median income.

I would suggest instead that we focus on two other problems with bootcamps, one mentioned in the article and one not:

* As the article suggests, bootcamps are probably not as effective at placing graduates as they'd like us to believe. There is a stigma to bootcamps, which is telling, because there's no such stigma for self-taught programmers. That may be because bootcamps aren't effective, or it may be because of correctable market inefficiencies that someone will make money arbitraging out. I'd guess it's a little of both.

* It's probably not easy to staff a bootcamp. They don't pay instructors particularly well. If you can effectively train 70% of a cohort of students well enough to secure professional employment, there's a lot of other stuff you can do in this field. To make the business work, bootcamps have to balance the willingness of students to pay with the salary demands of instructors, while retaining enough to grow. Maybe that's just hard to do.

I wouldn't move myself and my family to the SFBA for a $250k job. Perhaps that's cliché to say that kind of thing at this point. It needs to pay more due to the cost of living. I'm making > 200 now in a city-data.com cost of living index 91 city. SF is COL index 170.

The people who would take that job already have a similar job, more or less. It's not a skill gap so much as a "willing to put up with SF bullshit" gap.

Probably wise. I live in Daly City and my current cash salary pre-tax is just a bit south of $200k (ignoring nonsensical dollar-for-dollar comparisons to income like ISOs, performance grants and bonuses). I'm not struggling to make ends meet, but I am glad I have a big balance from prior my prior payouts, as I'm not saving much (it's even possible to go in the red for the month if I need to buy a non-trivial purchase over a few hundred dollars, like say replacement computer hardware or a major baby related purpose).

I know friends making 30-50% less who actually save more per month than I do, in Colorado. Now, the downside is they live in Colorado (compared to the weather and allergy situation on the coast, I prefer this life), but unless I do yet another startup they'll retire with more money saved on average per month than I have.

It's pretty fucked up that the entirety of SF life is basically priced around the few post-exit people who set the tone of the market. That is actually only a tiny number of people. Most younger devs I know save 0 money and just hope their wages will grow fast enough to offset that in the long run.

Oh and ironically I am certain I could get a job as a Haskell developer, though that's not my job. :)

Ok, I believe you, but what's your point? The only objective way to express a "skills gap" is the premium those skills command on the market. Meanwhile, it's obviously possible for there to be an outsized premium for a job you yourself wouldn't take.

We don't have to wade into the unproductive swamp of "how much it really costs a family to live in the Bay Area taking all quality of life factors into account".

Agreed, the salaries just don't scale with with the CoL in many cities. As someone in an index 93 city (Phoenix) I don't know how people are taking $100k jobs in SF and Seattle. $200k in an index 91 city is definitely great. Is that a dev position?

SF is 7x7 miles. Thats plenty of money if you're ok living elsewhere in the 6900 square miles of "bay area"

If I were thinking about working as an instructor at a bootcamp, low placement rates would probably discourage me from doing it - the last thing I want is to participate in a training program that cons jobseekers out of money, even if the salary is good.

amen to that. i took a look at one of those teacher gigs and couldn't believe how much they were charging the students. it seemed like i would be participating in a scam.

I have found when it comes to money, and career; 95% just go along with the flow. "Hay--its legal?"

(I got my first dose of reality in Chiropractic School. It was obviously placebo, but placebo was just not talked about. Selling out is so common. I'm not talking the cute TV version on of selling out, but the real people who con the uninformed, and vulnerable out of their money daily. There's a lot of denial in so many professions.)

You seem to have been shadowbanned for a long time now, but I didn't see anything in your comment history to justify it. Quite persistent of you to just continue posting! Maybe you should try contacting the mods to get your ban lifted?

FWIW: Two reasons (among many) that people can end up banned without anything in their comment record to explain it:

1. They've somehow participated in vote manipulation.

2. They were associated somehow with a sockpuppet account, which was itself banned.

It's probably better to stick to vouching for good comments that you think shouldn't be banned, and less on analyzing the reason for a ban (as you say, they can always just mail hn@ycombinator.com to find out what's going on).

If you effectively train 10% of the cohort, and some of that cohort find jobs, then several of those 10% would enthusiastically recommend you internally for any positions that come up in their companies.

or it may be because of correctable market inefficiencies that someone will make money arbitraging out

Talk to me about this in a couple years.

The article makes good points, but the author, nor anyone here really knows why they're going out of business, or even if it's a systemic thing. The industry is becoming more mature and competitive. It's inevitable that at this stage, some firms will fail and there will be consolidation. I know plenty of people who were able to get great jobs right after graduating from a bootcamp and quickly become valuable members of their teams.

Do bootcamps adequately prepare people to work as software engineers? No, but neither do most CS programs. I think the more pernicious problem they pose to their graduates is that they flood the hiring market will low quality talent. They don't do a good job of screening who will be any good, and just push them through their programs without ensuring that they know the material.

At the same time, employers aren't focusing on the things that really matter to them in the interview process. Making someone write a contrived algorithm on a whiteboard in under 30 minutes doesn't tell anyone anything. I get that companies need to search for a signal in all of the noise bootcamps are putting out, but they need a better way of doing that.

Economics 101 - because there are too many of them? There was demand; with demand, supply found its way into the market. As often happens, supply then overshot demand which led to lowered prices which led to shrinkage of supply.

... none of this will prevent a "Why are new coding bootcamps springing up like mushrooms after rain" piece of journalism appearing in a few weeks...

The two examples in the article were owned by large educational conglomerates: Apollo and Kaplan. They are clearer examples of big companies end of lifeing a product than examples of the Bootcamp market.

Since the salary a skilled and experienced developer can make in the current job market does not really align with University of Phoenix's adjunct pay scale and tuition rates, it is hard to see how a Bootcamp could generate similar rates of return to UoP's accredited college degree programs. UoP has a widget making process via the sharing economy similar to Uber's. The analogy might be if Uber recruited professional race car drivers for a service claiming to get riders to the airport faster.

Both Kaplan and Apollo have a different economic model than many bootcamps. Many bootcamps generate significant revenue from job placement in addition to or in lieu of tuition. Currently, the sweet spot of bootcamps looks like it might be outsourcing on the job training more than as traditional private business schools. Over the long run, programming will be better suited to the private business school model as programmers become common in small businesses in the way bookkeepers are.

Great teachers are expensive. This needs to be emphasized more. Even if there is strong demand for the dev bootcamps, the bootcamps still have to hire great teachers. The teachers are highly experienced developers, the kind of people who could be making something between $120,000 to $160,000. So the schools face some very high expenses.

A comparison might be made to retail. Sometimes you'll see a store that is very busy, full of customers, and then it goes out of business. Why did it fail when it was so busy? Often the reason is that the landlord raised the rent.

High costs can kill an otherwise good business idea.

What is the average boot camp instructor salary?

As a neighbor and local supporter of Iron Yard Tampa Bay (St. Petersburg FL), losing our growing economy's only code school was a big letdown this week.

In my experience, admittedly Iron Yard graduates were often too green to join our small engineering team (and we don't utilize .NET or Ruby). I'd like to say we have better training and mentorship for junior devs, but we don't yet. However, when we listed an iOS position, it was The Iron Yard's campus director who spread our ad through their national network. She found an Atlanta Iron Yard graduate who spent two years at a large company, receiving on the job training that now qualified her for our role.

The Iron Yard St. Pete was meaningful in that it provided a starting point for people in our area to make a career change and offered a meeting place for tech (located in the same building as our co-working space). The closure for us here locally is less about losing a code school and more about losing an anchor in our developing startup ecosystem.

Many bootcamps are terrible and promise things that aren't possible. (People generally can't learn how to code professionally in 12 weeks.) The bootcamps that tend to deliver do so because they only accept people who already have put significant work into programming before they start the bootcamp.

I organize a programming club in Berkeley that meets twice per week (over 4,000 members on meetup.com), so I meet a lot of people who go through various bootcamps. I think that I could help demonstrate how to fix many of the problems with the entire bootcamp model, if we could obtain funding for opening a physical space in Berkeley. I've been doing it for over three years and have a pretty good idea about how to find people who will succeed, without limiting opportunities or even charging a lot of money.

Maybe because too many people got into the business and completely saturated the market with coding bootcamps. I mean, maybe I'm wrong here, but I get the impression most people running these things didn't really do it because they were passionate about the field or knew how to run a teaching business, but because it was seen as a quick way to make extra cash.

There are just too many of the bloody things, and it's a pain to compete in the field as a result. Just look at how many this site seems to list for example:


426 schools listed there alone. And that's likely just the more promoted/credible ones.

It's like any type of business really. The more popularity a trend gets, the more competition there is, the more people who don't know what they're doing get involved and the more businesses/sites go out of business. It's happened to web hosts, to mobile app developers, to people trying to run social networking sites and marketing blogs alike.

The fact they're offering a service facing competition from both real world schools/colleges and online help sites probably isn't helping their chances either.

"than there are people looking to pay, on average, $11,000 for 12 weeks of intensive training"

...able to pay $11,000 and not work full time for 12 weeks


There isn’t really much evidence of a “skills gap”

Except that hiring is nearly everyone's biggest problem in tech and, you know, that whole crazy H1B thing.

Companies and Gov't job programs should be paying for dev schools not poor people. If they did it would all work out fine. As a matter of fact Iron Yard's corporate dev boot camp has been the profitable part of the business for some time.

I'm flabbergasted by this article. I'm not going to waste the time to reply to every one of the ridiculous assertions.

America desperately needs more skilled programmers in almost every market. Dev boot camps can and do get people to work. Many people in the position of needing a dev boot camp can't afford to go without work and pay for it at the same time (or at all).

> America desperately needs more skilled programmers in almost every market.

Outside of a few hotspots like SF, do offered wages really back that up? It's good being a developer but it certainly doesn't look like a desperate shortage.

In metro Boston area, we're paying new college grads $100K to start and sometimes losing on comp.

We're not "desperate", but for the overall market wage for new college grads to be 1.5x the state median household income, that smells like shortage to me.

The state median household income includes high school dropouts working at McDonald's, and it includes folks living in very affordable rural areas with low wages. It's not a meaningful point of comparison when you're looking at a subset of workers who are highly educated and seeking jobs in a technically demanding field at companies located in almost exclusively expensive metro areas.

To add to that, the Massachusetts median household income is nearly $70k. And in Boston? It's almost $80k. Moreover, the median family income in Boston is $100k.


Not specific to Massachusetts , but if you look at the BLS numbers, both the median income and job growth statistics are higher for Computer/IT professions than engineering/architectural professions.



You made sokoloff's point. A fresh college grad on day 1 of working makes the median family income in Boston.

To say that wages are not reflecting the shortage (the comment originally replied to), is ignoring the statistics.

Not at all. In fact, as far as I can tell, you completely ignored the point I made, so I'll just repeat it:

> The state median household income includes high school dropouts working at McDonald's, and it includes folks living in very affordable rural areas with low wages. It's not a meaningful point of comparison when you're looking at a subset of workers who are highly educated and seeking jobs in a technically demanding field at companies located in almost exclusively expensive metro areas.

Replace "state median household income" with "Boston median family income" if you wish. Many elements here still apply after that substitution and must be addressed in detail if you wish to claim that I "made sokoloff's point".

The part you replied to and didn't ignore was just a correction to incorrect statistics that had been cited without reference. It seems pretty warped to suggest that I made his/her point by indicating that the cited statistics were off by a factor of 2 -- and not in the direction that favors their argument.

I think it would also help if you defined precisely what you mean by "shortage". You seem to be using an extremely loose definition that claims a shortage exists any time new workers in a particular industry are paid above the median family income. We must then conclude that there is a shortage of all kinds of workers across countless industries. For instance, because new engineers in the O&G industry are typically paid $80k-$100k, there must be a shortage, even though the industry has been having layoffs for years and certainly does not struggle to fill needed positions.

> It seems pretty warped to suggest that I made his/her point by indicating that the cited statistics were off by a factor of 2 -- and not in the direction that favors their argument.

Which statistic did I cite that was off by a factor of 2? Your figure of $70,628 is 4% higher than the google search result featured snippet for "median household income massachusetts" of $67,846, but even using your figure, my saying that $100K is "1.5x the state median household income" isn't anywhere near wrong by a factor of 2.

1.5 x $70,628 is $105,942. My claim is at the very most wrong by less than 6%, not the 100% you seem to be claiming.

That's fair enough. You should really be using $80k though. It makes little sense to average over the entire state. You're then off by 20%, which is not really a negligible error. You're right, though -- I made a stupid error either way.

It doesn't really matter though because it's a deeply flawed comparison to begin with, as already discussed twice. This is all tangential to the points I was making and that both you and matwood have declined to respond to.

OK, if we take the Boston household figure or even the family figure, a 21-year-old SWE new college grad their first day on the job is making more than any of those.

That's evidence against a shortage?

I'm not replying to your other points, because I think I'm not understanding them.

The only point I was trying to make is that these general median statistics across the entire state or even Boston aren't controlling for enough factors. Programmers are generally well educated and the work is technical in nature. This does not describe a very significant portion of the individual workers rolled up in the general median statistics. I was trying to suggest you should be comparing to similar professions, such as engineering in general.

Programmers don't look like nearly such an outlier then. Would you say most engineering disciplines are experiencing shortages? I don't think that's true. O&G is one potential counterexample. Chemical engineers are still very highly paid, the big oil companies don't struggle to hire the best students from the best engineering schools, and yet there are also lots of layoffs going on. The layoffs and hiring freezes seem inconsistent with the notion of a shortage.

Law is another interesting example. There's no shortage of students graduating with law degrees. In fact, there's a vast excess, and a good chunk of them are unemployed or underemployed and will never work as an attorney. And yet many big law firms are still paying new associates big six figure salaries, and partners at major firms are still earning $250k-$700k.

I find this especially interesting and even counterintuitive because it suggests that if there is a shortage, throwing an endless supply of entry level talent at the industry won't necessarily solve anything.

I'm not sure there's anything that really needs "solving". I think there's a shortage (but not a desperate shortage) of programmers in the job market, at least a shortage of qualified ones.

Many of the factors you describe "well educated, work is technical in nature [presumably implying a requirement of intelligence on the worker" I see as rarity factors, increasing the market-clearing price for that work.

https://www.indeed.com/salaries/Entry-Level-Engineer-Salarie... has some stats on engineering job postings with salary. Overall engineering is eclipsed slightly by programming (replace "Engineer" with "Programmer"), but both have a fairly high starting salary range.

Oil and gas has a perennial boom/bust cycle that makes it an oddball. Everytime oil falls under $50/bbl or so, a lot of companies find themselves in trouble who expanded predicated on $60+/bbl oil. It's almost as reliable as the sun rising in the east that O&G will have more booms and busts in the future.

Why not just consult the BLS statistics? https://www.bls.gov/ooh/computer-and-information-technology/...

Moreover, there are plenty of remote positions, so a programmer's physical location does not determine what geographical market they're in.

Why not consult the BLS stats? They have the median at $79k, which certainly isn't bad but it's not much out of line with other skilled professions.

Again, I'm not saying it's bad to be a developer, only that I'd expect a truly desperate shortage to carry more of a premium. The figures still look like business or management careers are the way to go if you're looking for money.

Because becoming a developer, like any other discipline, is much more than than just learning to use tools (as quickly as possible).

Its probably that employers and students alike are starting to realize that expensive crash courses are often not the best investment.

I got the impression that The Iron Yard shut down because the company who acquired them made them. Everything else seemed to indicate that they were doing great.

The best bootcamps (in terms of placement rates with SV companies) are equivalent to 12 week long interviews that the interviewee pays for.

They screen for people who already know a good bit, and are really good at interviewing in general.

Then they put those people through an intensive 3 month long coding interview and place the people who make it through.

The problem with this business model is that there is a small supply of people that are capable of making it through, are willing/able to pay $10k+ to do it, and are willing/able to go 3 months without pay.

It's a joke to think that Joe JobChanger is really going to pick up all the skills and knowledge he needs to get a job at Google in 3 months.

Professional programmers with years of experience regularly spend that long just practicing white board problems to get try to get a job at Google.

The amount of students who can do that is just too small to build an industry around.

Parts of this sound like exactly what I was worried about when I first heard of the boot camps. An expensive training program producing a very limited skill set isn't really a sustainable model.

Coupling that with the for-profit schools and an investment environment looking for a big payout, this just sounds like a train wreck.

> Google’s director of education echoed this sentiment: “Our experience has found that most graduates from these programs are not quite prepared for software engineering roles at Google without additional training or previous programming roles in the industry.”

No shit? Of course 12 week programs won't help anyone get into Google. None of them advertise that. They're all web development or app development curriculums. Google is an aspirational position that requires a brilliant graduate with some great internship experience.

>None of them advertise that.

App Academy does. They list Google on their home page.

They do seem to place some people with Google, but from what I've learned looking into them though, I think it has more to do with selection than training.

To me, it seems like top bootcamps are 3 month long interviews that the interviewee pays for.

They aren't. Two specific coding bootcamps that 1) had been acquired by for-profit education conglomerates and 2) underwent relatively rapid expansion into second and third tier tech markets went out of business.

As evidence of a larger trend, this article cites a single quote from the CEO of "a private lender and an alternative accreditor for the fast-growing boot camp sector." This is unpersuasive.

If the skills gap/demand for 'coders' is correlated with the closing of two 'coding bootcamps' closing; does the author of this post also assume that some spontaneous, profound event will reverse our current trajectory? That h1b and other visa abuse by primarily tech companies will be permitted to continue, perhaps even expand? Will the direction of 'america first' soon make a 180 as well, perhaps the cost of outsourcing IT/programming will be reduced by imminent unprecedented legislation?

While I understand that the majority of our greatest thinkers can attribute the University of Phoenix to their sudden rise from obscurity, is their any sort of data available displaying how well they educate their students? I have seen far too many students graduate from university with computer science degree, with little to no understanding of the real world, or anything other than the concepts required to learn to program with java(exaggeration), but I can only imagine what sort of education another 'for-profit university' provides all the kids who want to become cool, hip, programmers and make 250k for browsing reddit remotely.

My guess is they spend most of their time/money figuring out how to locate and market to unemployed/underemployed Americans with access to student loans, and little certainty or direction in their lives; while spending zero time coming up with a metric via quiz/survey to determine good candidates for programming jobs, as that would likely reduce profits for said for-profit university.

I don't know much about these bootcamps, but I'm sure some of them do great things, however, I don't think the author of this post considered any of the likely relevant factors to the conclusion they are drawing about a possible trend here. Quality front page nothingburger rambling.

Those prices are too high. I dont think anyone has these amounts lying around unless they already have good jobs (and hence do not need this). These programs should compromise on immersion and cost about $2000 for three months.

I don't think someone able to teach the basics of programming would want to settle for $6,666.67 a month(assuming a class size of ten). That's also assuming they had no gaps or dropouts between classes. Sure, in your example it's no longer immersive but:

1. Immersion is what gets even the weaker programmers through. Non-immersion will drop the placement rate greatly.

2. There's still a lot of prep work so the teacher will still basically be a full time teacher. And that's assuming a one person operation. They'll need someone for advertising, outreach to local companies, and their own technical needs.

Teaching people in general isn't a very good business. Even the lauded head start programs that last much longer than boot camps and target kids (so they have both much more potential influence and more time to recoup the costs) achieve meagre results if any.

In fact, there are reasons to believe that human capital is mostly what you're born with plus (or maybe it's more of a multiplier) the accumulated achievements of your predecessors with very little input from organized education.



The demographics indicate that coding bootcamps are mostly retraining people who already educated and giving them a new set of specific skills. That's a good thing. Jobs change, and people have to learn new skills from time to time.

Applications are open for YC Summer 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact