Since I wrote a review of my experience with The Data Incubator bootcamp, a lot of people have reached out in one way or another wondering if they should bother with a data science bootcamp. Many people interpreted my review as very negative. To be clear, I do think there is value in the program, but also some pretty serious problems. Despite my ambivalence about the program itself, I have to say it was worth it for me and is probably worth it for many others.

The real value of Data Incubator (which I imagine goes for most data science bootcamps) was being given challenges that required me to learn, and being surrounded by people with the same goal. It’s much easier to learn that way than trying to learn on your own. It’s amazing how much you learn from just being around smart people trying to learn things. Having some direction from a course specifically designed to take you into the world of data science also gives you a good amount of breadth over the field of data science, which can be hard to get when trying to learn on your own.

However, I think there are some serious costs to these programs. They are stressful. They require a more-than-full-time commitment for a couple of months. Some may require you to move, though maybe temporarily (I moved to NYC for 2 months for Data Incubator). Some of them require you to pay up to $15,000 - Data Incubator offers a less competitive option of being a ‘scholar’, where you pay about that amount to attend the program.

I think the stress is probably worth it - it sucks during the program, but any quick way of learning a lot is going to involve a lot of stress. It’s hard to put a dollar amount that the programs are worth, so it’s hard to say if it’s worth taking the time off or even paying tuition. If you have financial support from somewhere, and you won’t be sunk financially if you don’t get a job immediately after finishing, it might be worth it.

The issue is, not everyone gets a job. These programs don’t work for everyone. It’s hard to know the rate at which attendees of a program get a job - the issue with most of the numbers these programs put out is they decide who to include in their calculation. A lot of students just lose contact with the program after losing hope of getting a job through them, and the program doesn’t count them when they tell you what percentage of students found a job through the program. This basically guarantees the program will be able to list a high success rate. So don’t trust the numbers they give (or at least ask more about if it includes everyone who enters the program, not just the people who ‘successfully complete’ it, since the definition of successfully complete may mean they either got a job or they stayed in contact with the program for 3 months after it was over). Regardless, if doing the program will put you on a financial knife-edge, it probably isn’t worth the risk.

There are alternatives to doing a bootcamp. If you’re motivated and can get a data analyst position that works closely with data scientists, you can transition from there. You can also take courses, do projects, and learn on your own in your spare time. I think these options are slower and probably harder than doing a bootcamp, but they can work, and are certainly safer. But if you feel like you're close - you have a PhD or relevant work experience and just need that extra boost to get over the edge - then bootcamps can be really powerful.