Allred-Austen_and Student_Joram_Clervius Austen Allred with Joram Clervius, one of his first Lambda School students. Photo: Lambda School

The way we interview and hire job applicants today is deeply flawed.

This is especially true in software engineering roles, in part because companies rely far too heavily on structured interviews that "test" candidates on a set of skills that isn't very much like the job of software engineering at all. Those tests often include whiteboard brainteasers and algorithms quizzes that aren't closely related to what an engineer does on the job.

As Laszlo Bock, former SVP  of Google's People Operations , describes in his book on finding the best employee, Google studied tens of thousands of its own hiring decisions and concluded its own interview processes (using algorithm quizzes and other tests typical in technical interviews) were broken. The findings showed the company would have hired just as many successful candidates by randomly picking them during interviews. If one of the most innovative companies in the world – with a massive budget – can't get it right, the system clearly isn't working.

Think of it this way: If you wanted to gauge whether somebody could play basketball, would you take them into a conference room and quiz them on some contrived gameplay scenario? Or would you meet them at an actual court and spend time playing basketball with them? Perhaps ideally, you would put them in a game scenario and simply watch them play. Obviously, the latter makes more sense, but for some reason software engineering hires are too often subjected to irrelevant, skewed or biased tests.

Recent data backs this up: Only 25% of technical interviewees are consistent in their performance. In fact, strong performers on the job "mess up" during technical interviews 22% of the time. 

It's a tall order, but I believe it's entirely possible for us to dismantle this old set of practices and build new ones that mitigate bias, align incentives and hire people who are truly successful, not just the ones who can game an interview. 

The need for this shift is even more urgent today. There's light at the end of the pandemic tunnel and, because of that, we're seeing a surge in hiring and increased demand for talent in software development (estimates expect a more than 20% job increase, with more than a million jobs already available). This is an opportunity for us to take a hard look at hiring and make real changes to our systems.

I'm the founder of Lambda School, a live + fully remote tech skills-based school. Since we're powered by Income Share Agreements (ISAs) – in which students don't have to pay tuition until they are earning $50K or more – our success hinges entirely on our ability to prepare and help our graduates land jobs. Recently, we decided to put our theory about the flaws in engineering hiring to the test by rolling out a new program called Lambda Fellows. 

Lambda Fellows is, essentially, offering companies a chance to take their interviewees to the basketball court. It's not unlike an internship, but it's designed to be less risky and more beneficial for both employer and prospective employee.

Here's what we do: We match hiring companies with a talented Lambda School graduate who will work on their team for a four-week fellowship. The employer pays nothing; Lambda School covers the Fellow's salary. At the end of the four weeks, if the company is impressed, it can offer the Fellow a full-time position. If the company decides it's not a long-term fit, they can part ways without financial loss and the graduate has gained valuable, paid work experience in their field. 

It's important that we protect our graduates' energy and time, however, so we require companies to meet a set of attributes before they can utilize the Fellows program. First, they must have an open software developer role (aka, there is a "real" opportunity for the Lambda grad to be hired), they must be willing to commit to giving the graduate a mentor for three to five hours per week, and they must come up with a suitable project for the graduate so they can measure their progress via real work. 

From day one of founding Lambda School, one of our core pillars has been incentive alignment between the school and the student. Lambda Fellows is just another step in this mission.

I believe so deeply in our students and graduates that we're willing to pay their salary to prove to companies how skilled they are. We started Fellows with a small group of a few dozen students last fall, and more than 70% of Fellows who completed a fellowship since then have received a job offer. In early March, I decided to tweet about the program to find more companies willing to get on board. We saw hundreds of responses, and dozens of companies are ramping up to start a fellowship. 

So, while it's too early to definitively say this is the exact model that will work forever, the demand we're seeing so far is very promising and I believe we are on the right track. For now, what we've learned from the demand we're seeing from hiring companies is that there are three core motivators for a program like this: 

  1. Confidence you're hiring the right person for the job. Recent data estimates that when companies replace a salaried employee, they lose six to nine months' salary on average. That's a massive cost. By creating a four-week, risk-free fellowship period, we're giving both the companies and the prospective employee confidence that this is the right fit long-term. Not only will they see how our Fellows write code and the quality of work they produce, but also how they collaborate in meetings and fit in with the team. 
  2. Ability to tap into new hiring pipelines. Old models of hiring have bias baked in, in part because credential requirements (like a four-year degree or traditional CS education) limit the pipeline of candidates. Today, individuals can gain the necessary skills in a variety of places, such as schools like Lambda, community colleges or nontraditional employment paths. At Lambda, for example, our students are based in all 50 states, come from a wide variety of life circumstances, and 33% of our students fall into underrepresented demographic groups, which is higher than the current tech industry average (we recognize this is a low bar, however, and are always striving to improve this). 
  3. Prospects who feel supported and engaged are more likely to succeed. We've seen success stories and have had many students hired from our original push in the fall. One graduate, Dominique Maack, was hired at Virtually. "Compared to any other university grad that I've ever interviewed, Dom knocked it out of the park," Virtually’s CEO, Ish Baid, told us. "We would've been silly not to keep working with her."

I've already received extremely positive and useful feedback on how Fellows are performing every day. I'm excited about what we're seeing so far, but we're not done. This is just one step on the path to better hiring practices and we have more work to do.

Companies that stick to the same exact methods for hiring for decades, without meaningful deviation, limit themselves and their prospects; we have an opportunity to change the way things are done. We've seen the data about why the old-fashioned way of interviewing isn't effective and why diversifying hiring pipelines is critical for both employees and companies – there are no more excuses. We now have the technologies and resources to right these wrongs and make a significant impact on the tech industry. 

(Austen Allred is the co-founder and CEO of Lambda School.)