How is a take-home assessment like a pineapple?
Much like how the internet is polarized on if pineapple belongs on a pizza, we are similarly divided on if take-home assessments belong in an interview process. A quick search will yield countless articles, posts, and professional advice on why take-home assessments are the best or the worst thing you could possibly add to your hiring process.
As with pineapple on pizza, the answer is a lot more complicated than “good” or “bad”. My goal in this post is not to convince you to add a take-home assessment to your process (or that you are virtuous for already having it) but to explain why you should add one if you should.
At its most basic, a take-home assessment is a step in an interview process where an applicant is given a chunk of work and asked to complete it on their own time. There are a number of potential benefits for take-homes over live coding interviews:
- Because they are asynchronous, take-homes have the potential to optimize your hiring team’s time by allowing them to review in batches.
- As the problem-solving environment is more similar to “on the job”, you can get a more accurate signal than a live coding interview. Just because someone is good on the spot, doesn’t mean they’ll be good at the desk.
- Many candidates know that live coding is a bad demonstration of their skills, and view these interviews as a “necessary evil”. Additionally many candidates we’ve talked to also hate the stress that comes with live coding and prefer the freedom to take their time
- With a take-home, you can present the candidate with domain-specific problems more easily than in a live environment, since these problems often require research and thought
- If a candidate is reluctant to do the take-home it could be a signal that your company is low on their list and that could save you a rejected job offer (although reluctance can also be a problem with the take-home itself, but we’ll get to that later).
We recently worked with a Series D company that was trying to improve their interview process. Their goal was to make their process more inclusive to candidates and less time-consuming for their team. We replaced their technical screening interview with a practical take-home assessment and saw great results. 83% of candidates preferred this assessment to a live interview. We also received comments in line with the points above:
- “Less stressful and more related to the job”
- “Challenges seemed more like learning opportunities than set backs”
- “It mirrors the work environment more closely”
Take-home assessments don’t come without risks though. The primary risks and hesitations we’ve seen are:
- Candidates are actually deterred by the take-home. This could happen if it looks like too much of a time sink with an uncertain reward. Many companies we talk to also fear that it’s the strongest candidates who are most likely to not want to do a take-home.
- Similarly, companies fear that if they are not investing time in the candidate, why would the candidate invest time in their process.
- It’s challenging to check for plagiarism when the candidate is completing the challenge asynchronously.
- There are far fewer resources out there for designing a take-home assessment than for running a Data Structures & Algorithms interview.
These are certainly real problems, but from our experience running take-home assessments, they are problems you can solve. For example, in the pilot project we conducted with that Series D company, 98% of candidates started the take-home assessment and 87% completed it, and the company was able to match their previous hiring rates with far more confidence in the applicants’ technical skills.
So how can you reap the benefits and avoid the pitfalls? Here are a few things to keep in mind when you are adding a take-home step to your interview process.
- Keep it short - Applicants may not have the time to (or may just not want to) sink an entire day into your take-home assessment. This can be especially bad for an applicant that is applying to multiple companies that all have their own bespoke take-home assessments. You also want to avoid making it so short that you can’t get any useful information from it, a good rule of thumb is under two hours. If your take-home needs to be longer (i.e. it’s a very involved problem, maybe for senior engineers rather than junior) consider moving it later in your process when applicants will have bought into your company and process more.
- Trim the fat - In line with the previous point, you should have a good reason for every problem the applicant is expected to solve. One example of this is providing them with starter code, since it’s quite rare for an engineer to be building an app from scratch on the job.
- Analyze and Iterate - Once you start giving your take-home to applicants, pay attention to completion rates, pass rates, what questions are being asked, etcetera and make changes accordingly. If you notice a lot of applicants are getting tripped up on one part (and it’s not what you’re testing them on), make that part clearer or remove it entirely.
- Compensate - If you are asking applicants to dedicate time at home to your process you should be offering them something in return. Whether this is by paying them for completing the take-home or simply providing useful feedback to them whether they pass or not.
- Connect - Take-homes have the potential to feel impersonal, to address this you can connect with each candidate to show the investment you would normally show in a synchronous interview. One great way to do this is to connect with the candidates over Slack so that they have a direct line to someone on your hiring team.
You can gain a lot of efficiency and information by adding a take-home assessment to your interview process, but only if you give your take-home assessment the same care you would give to any other step in your process.
So, why add a take-home assessment to your process? For the same reason you would add any other step, if you believe the information it provides is worth the effort of maintaining it.