A Bootcamp Bubble? August 01, 2017
There's been a lot of buzz lately about the recent high-profile closing of two coding schools: Dev Bootcamp and The Iron Yard. Depending on where you get your news, the back-to-back shuttering of these schools is either an anomaly in an otherwise healthy industry, or it's the first indication that the industry Dev Bootcamp pioneered is now ending.
Neither of these extremes totally captures the reality; the truth is more complicated. While I was certainly surprised to hear about these schools winding down, the challenges they described in their public statements definitely resonated.
In this post, I'd like to dig into one of the biggest challenges I've seen discussed. It's tied to one of Silicon Valley's favorite buzzwords, but is also frequently at odds with high-quality education. That word is scale.
Like many startups, scaling the business seems to be a concern for many coding schools. And in many industries, goals around scaling seem reasonable. After all, why start a business if not to grow it? If you want to benefit from market efficiencies only available at a certain size, scaling seems necessary.
But when it comes to education, here's a truth that many investors don't want to hear: high quality educational experiences don't scale. For bootcamps in particular, here are four reasons why:
Instructor quality and turnover. It's incredibly difficult to find and retain great teachers. While data on instructor effectiveness in coding schools doesn't really exist, a 2012 study by the RAND Corporation found that, at least in a traditional school environment, "Teachers matter more to student achievement than any other aspect of schooling."
Hiring great teachers is essential; however, retaining those teachers is just as important and just as difficult. This is especially true in the bootcamp industry, where instructors can often find more lucrative roles outside of education.
Real-world Experience. In spite of the difficulties attracting talent, a lot of coding schools have found really great instructors. Because of this, students can absolutely get a great education from an immersive program. I've seen plenty of students come out of my classroom and immediately begin a new career in technology.
All the same, aside from technical training, there are other skills you'll need as a software engineer that can be harder to pick up in a classroom setting. Understanding your team's programming and communication conventions, mastering more difficult git commands, or even reading through a large codebase are things you'll be doing frequently on the job, but rarely in a typical school setting. At Rithm, we're fortunate in that we've been able to partner with companies to give students experience with real-world projects in large codebases, but the only reason we can do this effectively is because we're small. These projects are great for students, but they're a huge investment, and take up basically all of our instructor bandwidth for one month out of every three. We've seen similar initiatives fail at larger schools precisely because the practice doesn't scale easily.
Student Pipeline. Aggressive scaling goals are often a high-risk, high-reward scenario. In the case of coding schools, growing to multiple locations or multiple course offerings means a significant amount of investment, which needs to be offset by increased enrollments. If you scale too quickly, this can result in a tension between doing what's best for the bottom line, versus what's best for the student and their experience.
Standardized Curriculum. Having a single set of curriculum that gets taught across campuses is certainly easier in terms of marketing a single product offering and training new instructors. However, it ignores the important reality that different markets prioritize different skills. If you're in Seattle, for example, getting a job in .NET will probably be easier than it is in the Bay Area. Even with technologies that are relatively popular, the degree of competition in a place like San Francisco will simply be higher than it might be in a smaller tech hub. This means that students who want to work in the Bay Area may need to focus more on computer science fundamentals to pass a technical screen.
Managing the nuanced differences in job markets across different locations is incredibly difficult to do well, and is another serious downside to scaling programs to multiple cities. It's certainly not an insurmountable problem, but seems like one that is often overlooked.
All of these have problems have one thing in common: they aren't problems that technology is particularly well-suited to solve. Not every problem is a technology problem. Or, as Audrey Watters put it in her recent piece on the state of the bootcamp industry, "'high-quality, immersive coding training' is going to be an expensive, labor-intensive proposition" (emphasis original).
I don't know myself to what extent these difficulties with scaling contributed to the closure of a few schools. Nevertheless, these are absolutely problems that we've recognized as we've built and run Rithm, and they're a big part of the reason why we're consciously not trying to scale our business right now.
It's a bummer to see these schools go. I'm sure the leaders of these organizations were trying to crack some of these difficult problems. Scaling a business is hard no matter what, but maybe some business aren't meant to be aggressively scaled. Maybe not every business should aspire to hockey stick growth. This is especially true in education, an area that is fundamentally about investing in people.
Written by Matt