I recently read the book Algorithms to Live By: The Computer Science of Human Decisions by Brian Christian and Tom Griffiths. This book discusses how we can apply algorithms to problems that we encounter every day. One of these problems is hiring, which is also known as the secretary problem. At its core, the secretary problem explains optimal stopping theory (when you should stop doing something to maximize results). In other words, when you should stop interviewing applicants to find your “perfect candidate.”
So when should you stop? When you’ve screened 37% of your candidates. While that may seem like a small number, it really isn’t.
Let’s say you interviewed 100 applicants for one of your open reqs. If your brain wasn’t fried by the end of it, your candidates would be extremely restless waiting to hear back for this one open position. If you follow the optimal stopping theory, however, you’d interview only the first 37 people to have the best chance of finding the “perfect candidate.”
Knowing the 37% rule allows you to optimize your pipeline and know when to stop recruiting and start hiring.
Let’s dive deeper. Imagine you’re hiring a full-stack engineer. You have 10 candidates in your pipeline and have scheduled them for an onsite interview. In your search for the perfect candidate, there are two ways you can fail. One is by stopping too early and the other one is stopping too late. So what is the optimal stopping point? In other words, when should you stop recruiting and offer a candidate the job?
If you hire the first engineer that walks through the door, you may leave the best applicant undiscovered. If you hire the last engineer that walks through the door, you have spent too much time looking and the best candidate was probably hired elsewhere. This is especially true in Silicon Valley, where competition for engineers is at an all-time high.
Therefore, the best way to go is by following the 37% rule. Hiring has a limited timeframe and limited candidates, so you need to make a decision quickly. In the above scenario with 10 candidates, the optimal stopping point is after the 3rd candidate.
The key takeaway of the 37% percent rule is not how many people you should interview, but how to optimize your pipeline. If you have 30 days to hire a new full-stack engineer and you can meet with 10 candidates, you should start hiring after the 3rd candidate.
Knowing that no matter how big the applicant pool is you will have a 37% chance of finding the perfect candidate saves you time and a lot of small talk!
About the Author:
My name is Cariny Perez and I am from Salinas, California. My hobbies include drinking wine, practicing Brazilian Jiu-Jitsu, and hanging out with my dog, Bear.