Issue #10: Mapping the bootcamp universe
Steal like an artist, map like an (amateur) geographer
The universe of bootcamp providers is vast. US-based non-profit Credential Engine this week released its 4th report on the state of the credentials market.
They found 1,076,358 different credential-awarding providers(!): from specialized certificates to bachelor's degrees to PhDs. If that number feels like a lot, you’re right.
Thanks for reading The View! Subscribe for free to receive new posts and support my work.
Non-academic providers – among which bootcamps and apprenticeships are classified – accounted for 656,505 of those certificates. That’s a lot of different ways to earn a non-degree certificate.
Data from the report indicates that there are 792 schools classified as bootcamps, of which 615 either have a physical presence in the U.S., are online, or both. Collectively, the online and U.S.-based schools offer 2,113 unique courses.
From just these numbers, it’s hard to visualize the spread of providers and how large they are compared to each other. A friend recently showed me Evan Shaprio's map of the media universe, and I thought it was a great way to visualize this type of data. I thought about how to create something similar for bootcamps and other accelerated training providers.
As Austin Kleon says, steal like an artist. So that's what I've done. I set out to create a version of this for bootcamps.
I looked at just US and UK bootcamps to start, and while I didn’t map all 792 schools, I made a start on a sub-section.
The first limitation I ran into is that revenue figures are hard to find. Most of the numbers in Evan's map are for publicly traded companies. Almost all bootcamp providers are privately held, and getting this level of detail is tricky without cultivating a number of sources inside each of these bootcamps. While I have some access, I don’t have Woodward and Bernstein on staff. I had to find another way.
Initially, I used various sources to calculate the annual revenue for each of the bootcamps on the map. I first looked at publicly available information (e.g., acquisition announcements). I created an average from industry sources (e.g., ZoomInfo, growjo, Owler, etc.), and finally I made a rough calculation based on the number of graduates and average bootcamp cost.
However, when I put the revenue figures on the map it felt like I was making far too many assumptions. And I didn’t want to get into a back-and-forth with any providers about the numbers. So I removed revenue from this map. I think you can draw your own assumptions about revenue if I share that the biggest players are likely doing $50m+ a year, mid-level around $30m and the longer-tail of smaller providers will be around the $10m mark or below.
I pulled expected graduates for 2022 from a variety of sources (Course Report, Track Changes, LinkedIn, outcomes reports by bootcamp) to create rough estimates for this year. These numbers should be taken with a pinch of salt as LinkedIn isn’t the most accurate and most bootcamps haven’t released their figures for 2021, let alone 2022.
There will be a wide swing either side for each of the numbers I’ve used. But, when taken in aggregate, these numbers are illustrative of likely graduates. And provide a good overview of how the market is put together.
Mapping the market in this way is a valuable way to zoom out and see some broad trends. What are my main takeaways from this exercise?
Well, I'm glad you asked; I have 3.
Diversified revenue. As I wrote in Issue #6, bootcamps that have achieved scale have two significant revenue streams. We can see this reflected in the map. The Big Four (GA, Flatiron, Le Wagon, and Ironhack) have robust enterprise and consumer businesses. The bootcamps that have smaller circles have only one main revenue stream or a much smaller second stream.
Emerging players. Microverse and Holberton School represent the only real innovation on this map currently. Their peer-to-peer learning model contrasts with the more traditional, lead instructor, 1:25 instructor model of the major bootcamps. We'll start seeing more innovation around the edges in 2023 and hopefully both Microverse and Holberton will release updated outcomes reports so we can see how successfully these schools are placing graduates into employment. I haven’t included any of the other bootcamp models I discussed in Issue #7 (employer-funded, hire-train-deploy, government-funded), I’ll look at this in a future map.
Long tail. Of the 600+ bootcamp providers, the most significant revenue spread is in the long tail of smaller bootcamps. These boot camps likely graduate 50% of the total bootcamp graduates each year, but they're hyper-local. Usually focused on one or two cities or graduating under 150 students a year. The recent Credential Engine report into the number of credentials made this even clearer. Bootcamp certificates have more than doubled from 2018 to today, with the category seeing a 38% growth since 2021. But the number of graduates has only increased by 13%. There might be more bootcamps, but there are fewer students to go around. And they’re being spread more evenly. As we can see in this handy visual:
What to make of all this?
This is most definitely a CAPITAL-LETTERS "Work In Progress," and I'll continue to tinker around with this to get up-to-date data each quarter to update it. My goal here is to offer you a new way of looking at market trends. Hopefully, this new map and the analysis that follows do just that.
Having put this together, there are a couple of things I’d like to do next. One: go back in time and map 2018 to today and create a nice visual to see how this landscape has changed. Two: map all non-degree workforce programs onto this map. At the moment, it’s heavily focused on bootcamps, but I’d like it to include a similar amount of credit-bearing or non-credit-bearing training programs whose primary purpose is to lead directly to work.
What are your takeaways? As always, hit reply to this email with any thoughts.
Take care out there.
Big missing piece for me is that 2U’s Trilogy “university bootcamps” are counted on the long tail instead of lumped together. They’re much bigger than GA in aggregate, afaict!
Simplilearn and Emeritus are smaller but still probably would show up if included.