Moneyball, the popular book and film, has become almost a cliche for businesses looking to hire talent on the cheap. Let’s take a look at how it applies to the problem of scaling coding bootcamps.
The 2011 baseball movie Moneyball told the story of how Billy Beane, general manager of the low-budget, second-rate Oakland A's, rebuilt the team after three of their best players departed for richer teams.
The A's couldn't afford to compete on recruiting with some of the other teams in their league. All the teams were competing for the same scarce resource. But Beane recognized that some players had characteristics that were undervalued. His insight? Finding how to identify these valuable but underrated commodities: a strategy that became known as “Moneyball”, thanks to the title of the book.
Using Moneyball, Beane helped the Oakland A's achieve a 20-game winning streak in 2002, while spending only $41 million on salaries. The team even competed against larger teams like the Yankees, who spent over $125 million in payroll that same season.
The problem the Oakland A's faced was the same as one of the problems that bootcamps face as they scale. As we saw in the last issue, there are 3 main reasons bootcamps fail to scale, one of which is that instructor recruitment doesn’t keep pace with growth.
The idea behind Moneyball, identifying talent that the market undervalues, is a strategy almost all bootcamps attempt to emulate when they reach a certain size.
Bootcamps have a tricky dilemma. They teach skills that are underrepresented in the market and that employers value, namely: data and software engineering. But finding people to teach these skills is hard because there aren’t that many of them – which is why bootcamps exist in the first place! You see the problem?
When they face this challenge, a lot of coding bootcamps look to lower their instructor costs. Instructor salaries are one of the highest costs a bootcamp incurs, so it makes sense to look at how to cut costs there.
What's a ready source of cheap instructors? Well, the graduates of your program. But without a process to identify top performers and train them, bootcamps will run into issues. You can’t just find an instructor, train them in a day and have them take on a cohort of students. Well, you can but you shouldn’t expect great results. For example, Lambda School incurred a lot of negative press when it was discovered their faculty was predominately made up of course graduates with minimal training, instead of experienced instructors.
Moneyball as a strategy feels like a good idea in this scenario, but, like all good ideas, the nuance of what the Oakland A’s did is often missed. Moneyball is a framework but it doesn’t work without a solid process. It's about more than just identifying under-valued talent; you need to have a rigorous approach to identifying the people who are the right fit for your organization, and also have the system in place to train them in the standards you are trying to set.
Too many people read these frameworks from famous companies and think that all they have to do is follow the framework. But there's no quick way to game the system. It doesn't matter what framework you use; the important thing is that everyone follows it, and you don't change it.
In January 2020, a UK bootcamp ran into this same problem. They solved it in two ways: the first was to have a framework, and the second was to have a process. One didn’t work without the other.
Let me explain.
A framework
When you ask most people who they want to hire, the typical answer will be something like "A players" or "only the best candidates". But you need a consistent approach to how you find these people. And what “good” looks like for me might be different to you.
Billy Beane and the Oakland A's understood this. They weren't just looking for cheap players. They were looking for undervalued players who fit into the team they were trying to build. They needed a good mix of players, and they needed to have a good coach in place to train them.
The bootcamp solved this problem with a hiring framework called "ICCE". ICCE stands for Intelligence, Character, Coachability and Experience.
At this bootcamp, the ICCE framework was rolled out to every manager in a series of trainings and, crucially, reinforced by leadership at each stage of the process. After each interview, managers had to submit detailed feedback on each candidate about why they felt they met each characteristic.
Many people in the company bristled against the idea that you could measure these things. But that missed the point. The point wasn't whether you agreed with the definition of 'intelligence' or 'character' that the company was setting. The point was that you needed to spot what the company thought intelligence meant, and then screen for those characteristics.
For example, intelligence wasn’t a candidate’s IQ or SAT scores. It was whether they asked insightful questions. Were they able to link concepts and ideas quickly and fluidly? Did they question assumptions?
Character wasn’t whether they had a good personality. It was whether they would enjoy the highs and lows of a fast-growing company. Were they coachable? Did they want to learn?
To test coachability, candidates would be given real-time interview feedback and assessed on how they responded. And track record: did they have a demonstrable history of success? Had they achieved something spectacular once (maybe it was luck?), or had they done something spectacular at every place they'd been? Could candidates identify their unique contribution to an outstanding initiative? Or had they ridden the coattails of other, more successful colleagues?
The ICCE framework was how this bootcamp ensured everyone who joined the company had met the same bar for excellence. It meant managers knew what to screen for in every interview, regardless of the role.
But a framework means nothing without a system to run it through. How did that work in practice?
A process
I'm an AFC Wimbledon season ticket holder. Wimbledon is a professional UK soccer team based in London. They play in League Two, which is the fourth rung of the English football pyramid. Bear with me; this is going somewhere!
The club is fan-owned, which means it is not backed by a wealthy financier or petro-dollars from abroad. Part of the club's ethos is they will always be fan-owned. But this means that they have less money to buy players than the other teams in their league.
Therefore, the club's recruitment strategy has had to incorporate youth players and, à la Moneyball, identify under-valued players from other clubs and youth academies. Their approach has been to invest heavily in the youth team academy at the club, and to bring in highly promising players who failed to make the grade at other London clubs.
The long-term strategy is to bring through these young players, train and invest in them, and then sell them. They can only compete with the other teams in the league by investing in these players now in the hopes of selling them for a profit further down the line.
In early 2020, the bootcamp we mentioned earlier had a similar problem to AFC Wimbledon. At the time, the bootcamp was a Series A company facing a massive scaling issue. They had recently launched a Data Analyst course, and demand was far greater than anticipated. At the end of 2019, they launched two cohorts. But there were 12 cohorts in the pipeline for the next quarter. Demand far outpaced the supply of coaches they had in place to deliver the program.
The bootcamp needed more highly trained coaches but didn’t have the resources to hire their ideal candidates. The target persona was an established mid-career data professional. The problem was that the bootcamp was competing with companies like Google, Facebook and other, more established bootcamps for the same people. Since they obviously couldn't compete with Google on attracting talent at scale, they needed to find people that these companies were not chasing after.
So when it came to solving the coaching problem at this bootcamp, the way AFC Wimbledon ran its youth setup was an example to follow. The bootcamp could only compete and hire data coaches by identifying underutilized talent and training them to meet its needs.
In March 2020, the bootcamp created a Data Academy to identify these undervalued, data-literate people. The target persona turned out to be recent science PhD graduates and graduates of data science bootcamps.
Both personas understood the statistics and data that the bootcamp needed in a coach. They were undervalued because they didn’t yet have the teaching expertise to be coaches. And they had yet to hit the job market. So they were typically open to taking on a role that paid less than an analyst who'd been in the industry for 1–2 years.
The ICCE framework took care of identifying the characteristics the bootcamp was looking for. And a 60-day teacher training program took care of their lack of teaching expertise (and sidestepped the issue that bootcamps ran into by hiring graduates and putting them straight into programs).
Still, the bootcamp needed a value proposition that was attractive to these candidates.
By offering them training in how to teach and be a good coach, the bootcamp could provide a skillset that would be valuable to future employers. And it had in place the system for these coaches to learn those skills on the job.
This became the model that helped the bootcamp solve the problem that Emma Rindlisbacher at Class Central identified as one of the 3 main reasons that bootcamps fail to scale: that instructor recruitment fails to keep pace with growth.
At least for the time being.
Footnote
I struggled to remember the 3 main reasons bootcamps fail until I started using the acronym “AHA!” (Acquisition, Hiring, Admissions). Now, when I think of the 3 reasons bootcamps fail to scale, I just think of seminal Norwegian synth-pop band A-ha or, and this one is a lot more specific, Alan Partridge.
Hey, it works for me.