Monday, September 06, 2021

A review of Lambda School from the father of a recent graduate

Background

I’ve been a professional developer for twenty years.  I exposed my son N to programming a couple times while he was growing up --  Scratch when he was around 8, Khan Academy javascript when he was 12.  He learned it easily enough but it didn’t grab him.

But his junior year in high school he had a hole in his schedule and I convinced him to try AP CS to fill it.  And this time, he got hooked.  He started programming for fun in the evenings.  You know how it goes.

Then in March 2020, Covid hit and his high school went virtual.  It was a terrible experience, to the point that instead of going back for more his senior year, he took the last classes he needed to graduate over the summer, and decided to apply to programming boot camps in the fall.  I think the American college system is broken, so I was happy to help evaluate his options for something different.

Evaluating boot camps

N and I came up with three criteria for evaluating boot camps.  If they didn’t meet these three, we weren’t interested:

  1. Income-sharing agreement, or similar.  The incentives for the school to, in effect, take your money and run are very very strong in for-profit education.  ISA means they only get paid if you get a job.  This creates a simple but powerful alignment.
  2. Modern curriculum.  If they’re still teaching Ruby on Rails, we’ll pass.
  3. Pre-work.  If a school will admit anyone who applies with a shotgun-style approach to student success, that’s not a good model.  It’s a much better sign if they have a rigorous set of pre-admission work to demonstrate some level of interest and aptitude.  I spent a little over a year teaching college level CS classes, so I know that for whatever reason programming just doesn’t fit everyone's brain.

Our shortlist was App Academy, Hack Reactor, Rithm, and Lambda School.  App Academy, Rithm, and Hack Reactor hit all three of the above criteria.  Lambda School did not have pre-work as rigorous as the others, but they offered the longest curriculum so we thought that could make up for it.  (However, Lambda changed from a 9 month course to 6 months just before N applied.)

While applying, we found that App Academy had some fine print that they would not do an ISA for students under 20, so we took them off the list. 

N was accepted to Rithm, Hack Reactor, and Lambda.  He decided on Lamba primarily because they have been online-only from the beginning so we thought they probably had an edge over Hack Reactor and Rithm, which had been primarily (HR) or entirely (Rithm) in-person before the pandemic.

Lamba School

Lambda (now renamed to Bloom Technology) did a pretty competent job preparing N to be an entry level web developer.  Enough HTML, CSS, Node.js, React, and PostgreSQL to be dangerous, plus the basics of Git and Bash.  A good foundation that he can build on.

Lambda School’s curriculum is a series of six month-long units.  After the first unit, most of the coursework was set up to prepare the students to tackle fairly meaty projects done in teams of five or six, divided into presentation layer, front end code (Node), and back end code (SQL).  These divisions are by experience level at Lambda.  So students X, Y, and Z would work on a project, then next month Z would graduate, X and Y would rotate positions, and W would join as the new guy.  Multiply this by two to get the six person teams.

The quality of instruction was overall solid, but if something went wrong like a version incompatibility with node, getting help troubleshooting depended on who you asked.

After the first four units came a month of CS subjects like binary trees and recursion, and then for the final month, they got back into teams for a capstone project.  N’s team revamped a web app for a tiny nonprofit.  It was good experience, he learned a lot about understanding what an existing system did and how to rebuild it while keeping the good parts.

So on balance: while I totally get the standpoint that there is high quality instructional and reference material across the Internet for free or a much lower cost than Lambda School, I think that between the actual instruction, the accountability from a formal curriculum, the project work, and the real-world capstone with the nonprofit, Lambda delivered value for what N is paying.

Post graduation

So the coursework and instruction at Lambda was well done.  Unfortunately, Lambda came up short in several areas of helping N find a job.

  1. Before finishing the coursework, there was little communication on how the job application process would go, what to expect, or how to prioritize your time.  
  2. Resume building was a mixed bag.  They did help N create a plaintext resume, and they coached the students on how to present their prior, non-programming experience, but they did not help create a rich text version for human reviewers.
  3. N doesn’t know how you get picked for one of the sexy new programs like Lambda Fellows, nobody talked about that and nobody he knows was included.
  4. N graduated right in the middle of traditional summer intern season, which Lambda ignored completely.  Seems like a missed opportunity at the very least -- surely most of the Lambda grads would prefer a paid internship to months of applying to developer positions while working another job.
  5. Most importantly: after graduating, Lambda emailed N just five job openings over a period of two months to say, contact them here if you’re interested. That’s it, that’s the extent of their post-graduation job search support. (None of the five replied to N’s application.)

On the positive side, N thinks Lambda had a great program for interview coaching. They did multiple rounds of one-on-one mock interviews to help the students get used to the kinds of questions they could expect in the interview process.

N ended up finding a job through my network, several of whom were willing to interview a new boot camp grad.  (Thank you!)  He just finished his first week working full time as a software developer.

Commentary and educated guesses

Given that N thought (and I think, and the people who interviewed him thought) that Lambda’s actual instruction was good, why are there so many reviews online complaining about it?  I think there are two big factors:

First, Lambda is doing something new and consciously not following traditional instructional design because "that’s how we’ve always done it."  Remember, Lambda was online-only before the pandemic.  That alone means things are going to be different from other schools.  And they’re not trying to help you get a well rounded classical liberal arts education or even necessarily to “learn to learn” -- their goal is to teach you enough practical programming to get a job as an entry level developer.  This means, for instance, that they do a lot more project work in teams than your nearest college CS department would.  I think that’s a good thing.

It also means that nothing is sacred and things can change quickly.  So they changed it from 9 months to 6 (which I believe did not affect already-enrolled students) and eliminated paid team leads from their project work (which did).  If you try new things, some of them aren’t going to work out.  I understand how this would suck as a student, but running a school is expensive, running a school that does something nobody has done before  (successfully, and at scale) is even more expensive, so the faster they can iterate on what works and stop what doesn’t, the better.  I don’t fault Lambda for this.

The other factor is students who didn’t have the necessary background to be successful.  It makes me sad to see Lambda students writing about “flexing” (repeating) a unit for a second time.  I think there’s a high likelihood that they weren’t ready to be admitted.  This is something Lambda could fix by increasing the rigor of their relatively short precourse work.

It’s a balance -- you don’t want to only admit students who are 100% guaranteed to succeed, but on the other hand it’s not really doing people a favor to admit them if they only have a 10% chance.  I’m not sure exactly where the balance is, but it seems likely (based on what I see other bootcamps doing that create high-quality outcomes) that Lambda’s filter should be a bit tighter.  (On the other hand, I see other people criticizing Lambda for making money off the students who are so well prepared that they would be able to get a job programming no matter what they did.  This is definitely not the case for the typical Lambda student, but if people are saying that then maybe that’s an indicator that Lambda has about the right balance after all.)

Based on Lambda’s relative lack of help sourcing job opportunities for N, I also wonder if they’ve scaled too fast, too quickly.  Lambda advertises two things: relevant skills, and help finding a job.  It seems to me that it’s a lot easier to scale the instructional part, than your pipeline of companies who want to hire graduates from a new and relatively unproven school.  This would explain the relative lack of referrals that N saw, and it would also explain why Lambda hasn’t released student success metrics since 1H 2019 over a year ago.  (And for that, I do fault Lambda.)

TLDR

Lambda did a good job with curriculum and instructional design, maybe even a great job.  But their job-search program was significantly weaker, or perhaps it just hasn’t been able to scale to meet an increased volume of admissions.  I am cheering for Lambda and I hope they can fix it.