The Problem

A lot of times when a start-up starts to build out their company, they hire smart people to work on problems, and this typically works well — to a point. At some point, a group of successful entrepreneurs is going to have to hire someone outside of their field and scope of expertise.

This typically includes things like office managers, marketing, sales. In addition to this, there are other aspects, like hiring specialized engineers — like systems administrators, network engineers, and backend engineers. At some point, these scopes become specialized enough, and critical enough that, realistically, the contingent of interviewers available at a company are not capable of properly assessing the technical skills of the candidate. Though, they are capable of interviewing the candidate on culture, and fit scale

In addition to this, fairness is a major aspect of interviews. With a growing diversity problem in the tech industry, properly hosting interviews to make candidates feel comfortable with their surroundings is key to building a diverse team. Building an interview process without bias is incredibly hard for existing organizations. Questions regularly come up around fairness, bias, and diversity.

The Solution?

Interview as a Service. I want a new model for recruiting, or rather a new step in the recruiting pipeline. I want this service to be attuned to the needs of the company. I want to use the most modern interviewing techniques, and ask questions that are relevant today. I want to remove the tax on engineers for interviewing. The difficulty of interviewing was well summed up:

There is this crazy mythos around here that everyone should develop their own interviewing style from scratch. In spite of having many strong opinions about hiring, I totally reject that. I unabashedly want somebody to spoon feed me proven best practices and let me follow them. The pipeline is too dry to risk losing good engineers on split testing, and it has a terrible feedback loop.

I’ve been really good at interviewing, but keeping up with changes (e.g. started getting candidates who had obviously studied my favorite algorithms question) takes too much time away from shipping features.

-Steven Lumos

The benefits are clear to me of such a service —

  1. Higher capacity / Lower latency to the interview pipeline: In today’s world, time to offer is key in the recruiting process. IaaS (Interview as a Service) can act as a method to quickly to screen candidates on their technical merit without interrupting your engineers. Leave the aspects that matter to your engineering organization — to your engineering organization — Culture.

Steven points out that staying on top of current questions in interviews is incredibly difficult. There are people who’s full-time job it is to create interviews; it’s incredibly distracting for an engineer to stay up to date on these methods. InterviewCake, a popular website for interview practice questions has 42 questions today. Project Euler on the other hand has 500 problems at this point. Both of these platforms are actively either adding or changing their corpus of questions.

There are people who’s full-time job it is to create interviews; it’s incredibly distracting for an engineer to stay up to date on these methods

Not only do questions that need to ask for specific positions change over time, but often times there is a requirement to hire a specialist over a generalist. Even with the advent of the “Full-Stack Engineer” we still need sometimes needs experts in networking, databases, or frontend technology. Acquiring these experts is sometimes a daunting task, even if the work is well known.

The interviews that I’ve experienced in the technology industry are a mix between behavioral interviews, technology design & architecture, and whiteboard-coding interviews. I choose not to lump the latter two together, as often times, whiteboard-coding interviews are seen as a problem solving exercise, whereas design interviews are seen as a way to gauge subject-matter expertise. These traditions find themselves constantly find themselves open to criticism, but are rarely put under a microscope. There are many voice in the community that argue for and against these, but neither has been put to rigorous, academic testing.

Call to Action

Let’s make interviewing an actual first class skill-set. I’m not saying you should outsource all your interviewing, but I think it should be taken as more important than it is today.

Let’s stop stressing out resident engineers from having to keep up to date on methods, and mechanisms for interviews, and outsource things that are not core competencies to external organizations. Recently, there’s been a trend to outsource more, and more critical infrastructure, and that’s the only purpose problem-solving, and basic skills interviews serve.

Let’s stop wasting the time of interviewees, and stop forcing them to continuously prove their skill set. There are known unknowns that occur across almost all interviews, and we can amortize those across all potential interviewers. The arguments against whiteboard coding are well thought out, and it discriminates against a particular type of thinker, and person, but due to the time-constraints of traditional interviews, these are the go-to mechanism for proving skills. Removing that constraint will only serve the industry well.

Let’s stop wasting the time of recruiters. Enabling recruiters to quickly survey the field for particular specialist skill-sets would greatly pare down the top of the funnel. We can remove the ephemeral nature of skills on resumes. This will help candidates flow through the pipeline more quickly.

What I’m advocating for isn’t outsourcing the entire interview process. I’m not advocating for bringing a system like the SAT & College Board into industry, but something closer to Hacker Rank. Let’s enable an opt-in model to allow for greater efficiency, and opportunities.