Lambda School Pt. 1: Consumers


Here begins the series of posts analyzing Lambda School, as outlined in my last post. By way of introduction, Lambda School is a Y-Combinator-backed startup that aims to upskill untapped talent to participate in the tech industry. They’ve raised ~$4M and are currently offering intensive courses on computer science, data science, iOS development, Android development, and user experience design. I’ll address these pieces in more detail during the “Company” post in this series.

A Screen Cap of Lambda School’s Website

With that intro in mind, the purpose of this series is to understand the strategic choices Lambda has made within the frame of larger market-based and societal considerations; it is also to determine what Lambda could do in the future and/or how Lambda’s choices could be co-opted into other businesses, sectors, or disciplines.

Disclaimer: some of this analysis may be deeply flawed as I’m relying only on public information and my own lived experience. I have no access to the guts of the company as I used to with my consulting clients, so a lot of what I do here will be conjecture and opinion. This will result in the construction of narratives that are most likely only partial reflections of Lambda’s reality. I suspect that how correct I am will follow a normal distribution: mostly somewhat right with a few strong hits and complete misses.

The Purpose of Higher Education:

Before jumping headfirst into a discussion of Lambda’s consumers, it’s important to first understand what higher education itself is all about. While we’ll do a deeper dive on this in the “Category” post, let’s quickly look at why consumers pursue education.

According to Jeffrey Selingo of the Washington Post, “college was once seen as a place where adolescents went to explore courses and majors before settling on a job and career, often well after graduating.” It was a place in which knowledge was pursued for its own sake (Stanford). 

But all of that changed on Feb. 28, 1967, when Ronald Reagan, as governor of California, proclaimed that taxpayers shouldn’t be “subsidizing intellectual curiosity” (The Chronicle of Higher Education). Since then, higher education has become a means to an end: a way of getting a job.

While I’ll save broader discussion of education as a sector for later, there is one takeaway from the preceding paragraphs that’s critical for today’s post: the shift in higher education from learning-focused to employment-focused has fundamentally transformed the nature of education. College has transformed from a place to expand minds into a marketplace for talent. Its two sides are employers and students, which is what we’ll focus on today.

Consumer Group 1: Employers

Employers are in dire need of computer scientists, a conclusion that the data bears out in an obvious way. “Using statistics from the Bureau of Labor Statistics (BLS), it’s projected that 1.4 million positions will be open in computing with only 400,000 computer science grads. Hence, there will be a shortage of 1 million programmers” (TechCrunch).


However, claiming there’s a talent gap based only on CS graduates is foolhardy. Data shows that 60% of software engineers don’t have CS-specific degrees and that 36% of IT workers have no degree at all (TechCrunch). Given that software engineering doesn’t have any licensing requirements like accounting, law, finance, or medicine, it makes sense that this industry can absorb atypical talent.

Even so, employers complain about their inability to find talent: 55% of executives state that inability to access developer talent will constrain their company’s growth (Stripe). Yet 56% of those same execs say the number of developers has increased. How can we square the increasing number of developers with employers’ inability to find them? 

One reason might be the exorbitant salaries paid by the GAFA companies (Google, Apple, Facebook, Amazon). The median wage at Facebook is above $240k (Recode), which would certainly draw talent away from other companies. But I don’t think that’s the problem. Even if they wanted to, GAFA couldn’t absorb all the developer talent out there for a number of reasons: some people don’t want to move to major metro areas, some people have ethical qualms with GAFA, some people are uninterested in GAFA’s projects or feel limited by these companies’ bureaucracies, etc.

The reason companies complain about the difficulty of accessing talent is more likely the fact that there’s no way to separate signal from noise. Hiring is a difficult and time-intensive process. The best way to make it more efficient as a hiring manager is to go for the low-hanging fruit (i.e. optimize for factors that scream “this candidate will be a fit and is worth your time”). That’s why tech companies go to recruit in college computer science departments: the kids are bundled into a nice pool for the companies to sort through via the CS department and are effectively pre-certified by virtue of their degree.

Conversely, it’s inefficient for employers to sift through holders of other degrees or non-degree holders. There’s simply no way to tell, at a glance, that someone has the requisite skills. Because time allocation is a zero-sum game, companies choose to spend their time on high signal sources that will more likely produce ROI. This is how employers are consumers of higher education; they rely on colleges to bundle up prospective talent and certify their skills via degrees, making the talent acquisition process much easier for them.

And yet, this over-reliance on colleges to do the filtering is what has artificially created the appearance of a talent gap. It’s relatively difficult for  non-CS degree holders (graduates or otherwise) to send enough signal to employers to pique their interest. As Vivek Ravisankar, CEO of HackerRank, puts it: “software leaders’ deep-seated fear of hiring anyone that will slow them down is systematically locking out millions of skilled engineers who lack a high pedigree while forging a so-called talent shortage” (TechCrunch). Ultimately, if the gap in this labor market persists only due to employer hyper-focus on pedigree, the solution is finding ways to enfranchise atypical candidates.

Consumer Group 2: Students

With the employer consumer group in mind, let’s consider the other side of the equation: students. Who is the sort of person who would want to study at Lambda School?

While I’d like to offer some hypothetical psychographic personas, I’m reluctant to do so without real data from Lambda, so I’ll hold off on that for now. Instead, I’ll try to focus on some overarching demographics and psychographics, along with the draw of Lambda for such students. Disclaimer: again, these are all guesstimates because I don’t have official data.

Based on what I’ve seen, Lambda has a much wider spread of demographic diversity than any academic department or corporation I’ve ever seen. While the student body does seem majority Caucasian, I don’t think it’s a huge majority. It’s not uncommon to find people of all sorts of races and ethnicities, which I think is an upside in a learning environment, as it forces you to deal with people with hugely different psychographic frames. The age spread is much much wider than what you’d see anywhere else, I think. There are lots of folks over 40, in contrast to the cliff that you might see in Silicon Valley employees at around that age. On the other hand, I don’t think there are any 18-year-olds, like you’d see in a college.

It appears there’s a mix of degree holders and not. Non-degree holders seemed to lack them for two main reasons: (a) tuition was too expensive to be worth it or (b) unexpected life circumstances forced them to leave college. For degree holders, it seems that students’ choices of major didn’t necessarily lead to robust career opportunities or that systemic change (e.g. automation, outsourcing) in their industry of choice caused them to reconsider their options. 

As an aside, the growing disconnect between most degree programs and employment opportunities has recently led to a notable reduction in humanities degrees and growth in more “employable” fields like the sciences.

Based on one-off posts shared by Lambda about how much graduates had increased their income, I think it’s safe to say that most Lambda students made around or under the US median income (~$45k) before starting the program.

Hence, I think the primary drive for students to learn at Lambda appears to be career and wage improvement at no upfront cost (Lambda offers conditional tuition). Given the prominence of technology even in non-industry media, I’d argue that most people know at least a bit about what’s happening in the industry and how much money engineers can earn. I think tech’s outsized role in media narratives has created an anchoring effect around the possible opportunities in the space, motivating people to try to break in.

However, I think there’s also a strong subordinate motive for many students: to problem-solve and build things. In the student Slack, people are extraordinarily excited and proud to show off things they’ve built using the skills they’ve gained. There’s also a tendency to help peers debug technical issues and troubleshoot problems. In CS, there’s very low barrier to creating and releasing projects. Something tangible that you can feel proud of as a learner (e.g. a simple website) is not super hard to create, whereas in disciplines like biology or English, meaningful progress is capital- or time-intensive to the point where it can feel like you’re not moving along at all.

Intriguingly, many Lambda students use language related to “change” and “transformation” to describe their experience, as if attending the school is like a metamorphosis. You enter in one state, and leave completely changed. The optimism and excitement expressed by Lambda School grads is astonishing; most people I meet don’t report that their education fundamentally changed their lives.

But when you look at Lambda School outcomes, you’ll see that students are often doubling or tripling their pre-Lambda income. These sorts of results are life-changing for students. Notably, they will also serve as sentiment drivers for the school in the future. I expect that customer satisfaction and net promoter score metrics for Lambda are or will soon be extremely high.

With all of this in mind, there are some immediate challenges in Lambda’s way. First, they will need to get over the perception that Lambda is “too good to be true.” Based on student review sites, some students said Lambda initially felt scammy due to the conditional tuition model until they were able to verify its legitimacy through social proof. Conditional tuition is a huge differentiator, but there’s some work to be done to convince people it’s not a scam. One way to do that might be written and video student testimonials in which students acknowledge potential fears and explain them away in their own words.  This might be coupled with access to select student ambassadors on Twitter and LinkedIn who can explain the tuition model to prospective students (as a form of social proof).

Second, Lambda might need to get over the widespread perception of “tech as magic.” What I mean by this is that the inner workings of technology are an arcane topic to many. If I asked you how your computer or the internet worked, I doubt you’d be able to explain; I certainly wouldn’t in any meaningful detail. The fact that tech is complex causes us to overestimate the complexity of tech jobs. What people don’t seem to realize is that most jobs are actually just combinations of easy skills bundled together. For example, consulting is just speed-reading, strong writing, and an above-average ability to use PowerPoint. Investment banking is public speaking and Excel. Software is really just algebra and logic. Framing tech as a combination of easy skills might help prospective students overcome silly fears and get over impostor syndrome.

In any case, Lambda is clearly focusing its efforts on the atypical candidates mentioned in the previous section. The main area of improvement might be increasing the size of that funnel by handling student objections quickly (as in the above 2 paragraphs).

Lambda’s Strategic Consumer Choice:

Ultimately, strategy is about decisive choices. Truly differentiated strategies are always about non-obvious, non-stupid choices, which I described in my post about contrarian questions. So today, we’ll try to understand the choice that Lambda has made with regard to consumers.

On the employer side, Lambda has realized that the false shortage of developers can be corrected: there are lots of folks who have the right mindset and inclinations who just need some upskilling and a nudge in the right direction. By enfranchising atypical candidates like non-degree holders, Lambda is effectively arbitraging talent from places they’d never be discovered to the firms that need them.

Eric Koester, the founder of Zaarly, says that you can either borrow signal or make signal. The former refers to establishing credibility through external mechanisms (e.g. going to Harvard, working at McKinsey). The latter is about doing work that establishes you as an expert (writing a book, building software, starting a company). Lambda School does both for its students. The Lambda name acts as a certification for graduates, conferring signal upon them. But all of the projects completed during Lambda, including open source contributions and personal projects, involve students making their own signal.

Ultimately, this high signal arbitrage of talent provides a way to solve the talent discovery problem for more obscure companies (i.e. non-GAFA, non-unicorn), but still have high willingness to pay for talent. In fact, Lambda may even be more attractive to employers because the Lambda talent pool enables them to forego competing for a highly in-demand, highly limited group of CS majors.

On the student side, Lambda is giving hungry folks a chance at success. While we’ll get into their tuition model during the “Company” post, it’ll suffice here to say that payment is contingent on a given student getting a well-paid job in tech. This enables disadvantaged students to make a leap in their career without paying an arm and a leg upfront.

But more importantly, Lambda assists students in creating high slope to make up for their lack of y-intercept. John Ousterhout, a Stanford professor, argues that y-intercept refers to your starting position in life (i.e. privilege and pedigree) and slope refers to the speed and volume at which you can learn and attempt new things. Just as in math, someone with a low y-intercept and high slope will eventually surpass someone with high y-intercept but low slope.

By condensing the equivalent of a 4-year CS degree into 30 weeks, Lambda is putting students on a high slope trajectory, forcing them to grapple with new concepts at a breakneck pace. This ultimately develops good learning habits along with an ability to handle cognitive adversity. It also proves to employers that the student is a fast learner. 

With this in mind, it’s clear that Lambda’s strategic choice with regard to its consumers has been highly cognizant of the dynamics of the 2-sided market laid out above. If I were to outline their choice in a tweet, it’d be something like this:

Enable disenfranchised students to gain CS signal in the eyes of employers through high-velocity project-based learning while simultaneously breaking the employer bottleneck of talent discovery.

Per my contrarian questions post, this is non-obvious and non-stupid. In fact, a prospective investor turned down Lambda’s founders, claiming it was silly to try to educate the “dregs of society” (mind you, that’s a stupid take; Lambda students are really smart, but then again, I’m biased). But if sophisticated people can vehemently disagree about a strategic choice, it’s probably on the right track of being non-obvious and non-stupid.

Future Strategic Implications

The biggest thing I suspect Lambda can do with regard to consumers is to build out more robust employer-focused products. Lambda currently offers Concierge, which is effectively a free placement service, and Apprenticeships, which selects and trains students for a given employer’s tech stack in exchange for a paid internship for that student.

My hope is that these offerings will one day evolve into a fully-fledged staffing and placement division. That enables Lambda to create another revenue stream (many placement firms charge 20% of the candidate’s salary upon successful placement). This could potentially lower student tuition, opening up the admissions funnel for even more candidates, who may be wary of the conditional tuition model. 

Source: Top Echelon

I’ll leave off here for now, as this post has gotten quite long. As always, please feel free to leave me any comments or questions you may have. 

A Taste of Things to Come…

For my next few blog posts, I want to try something new: writing a series. I’m imagining probably 4 somewhat interconnected posts in the service of telling a bigger story or making more nuanced points than I usually do. I feel that since my return to the blog, I’ve been missing some of the deeper…

Post #100: The Genre of Progress Literature

This is post #100 on Thought Distiller! I’m personally surprised; I had no idea the blog was so far along in terms of post count. Anyway, thanks for being a reader! Here’s a post that, by the end, I hope will lift your spirits. What’s wrong with the news? It often feels as if we’re…

Conflating Poverty Alleviation with Income Inequality Reduction

I’ve often found that writers— even those trained in economics— conflate alleviating poverty with trying to make the societal income distribution more equal. The most recent example of this that I’ve encountered was within Rutger Bregman’s acclaimed book Utopia for Realists. I enjoyed the majority of the book, which advocated a universal basic income (UBI), a…

The Business of “Fake” Reality

More and more, it seems like the word “fake” is cropping up in our public discourse. Whether that refers to Trump’s favorite phrase “fake news” or uncanny and disturbing “deepfakes,” fakeness is rapidly pervading various forms of multimedia. In the past, people could certainly have created work anonymously, pseudonymously, or even forge work under someone else’s…

And We’re Back!

It’s been quite a long time (7 months!) since I last posted to Thought Distiller. My hiatus occurred due to a confluence of factors that prevented me from devoting the time and attention that I would have liked to this blog. And I realized that I would rather write nothing at all than to write…

Thinking about Apps Inside of VR

In late November of last year, I tweeted a bite-sized view of what I think the future of VR (I’ll use the term VR as inclusive of AR) will look like: The point of this post is to expand a bit more on why I believe this to be true and what needs to happen…

Six Links: Pirates, Moats, Capitalism

Hopefully this edition of Six Links will help you find yet more intriguing content to help you finish January strong. Here goes: The Hijacking of the Brillante Virtuouso by Kit Chellel & Matthew Campbell: Anytime a commercial vessel is lost, the incident is recorded with a quill pen in a leatherbound book at Lloyd’s, a London…

10 Predictions for 2018

I saw a number of prediction posts for the upcoming year as I was surfing the web today, and I figured I’d write my own. So without further ado, here are ten predictions for this year: Donald Trump will still be president at the end of 2018 (confidence level: 8/10) The cryptocurrency space will attract…

A Brief Exploration of the Cute

In 2016, I wrote about what makes things “creepy” in another brief exploration post, which you can find here. This time around, I’d like to take a look at what makes things “cute.” Interestingly, but unsurprisingly, our notion of cuteness stems from evolutionary biology. In 1949, Nobel laureates Konrad Lorenz and Niko Tinbergen described the “Kindchenschema (infant…