Have you ever wondered what the interviewing process looks like from the other side of the curtain? Maybe you’ve interviewed for a position and thought you did really well, only to receive a rejection with no explanation as to why. Or maybe you’ve absolutely bombed in an interview and then, a week later, received an unexpected offer.
Today, I want to talk about the hiring process in big tech. That’s Facebook, Amazon, Apple, Netflix, and Google, otherwise known as FAANG (as if you didn’t already know).
The FAANG Hiring Process
You probably already know (or at least suspect) that big tech’s hiring algorithms are extremely methodical. These are companies with multi-billion dollar valuations that have to operate at scale. That means there are lots of gaps in their operations even in the best case scenario. And they still have to try to find the best hire for the job.
Maybe you’ve even interviewed at a big tech company. Whether you were rejected or given an offer, you may wonder what went into that decision. Today, I’m going to focus on Facebook.
Because Facebook in particular is extremely careful about introducing minimal bias into their interview process.
The Facebook Hiring Process: A Case Study
In addition to basic competence, Facebook interviewers are known to test for:
- Fit: Viewing the candidate as a potential colleague, including their ability to understand complex ideas, maintain enthusiasm and motivation, and make impactful decisions quickly
- Generalism: The ability to be flexible in terms of the role you’re playing in the company’s structure, to move from one domain of expertise to another in a pinch
- Architecture: the ability to visualize a problem and solution space and approach problems with unusual constraints
- Coding: usually done on a whiteboard, problems are explained in a few minutes and expected to be solved within half an hour
The Facebook onsite interview is a product of the opinions of seven to eight different interviewers. Facebook requires its interviewers to remain completely stoic to reduce the effect of their reactions on the candidate. Each interviewer has their own area of expertise and submits a report with two components: their decision (“hire” or “no hire”) and their confidence in that decision (ranked from “Absolutely confident” to “Not very confident”).
Here’s the first bottleneck: in order to even qualify for candidate review, the ratio of “hire” to “no hire” decisions has to work in your favor. The recruiter rejects the “bad” candidates and forwards the “good” candidates to the candidate review process.
The Facebook Candidate Review Process
The recruiter then sorts the reports for the individual into “packets” that she forwards to the candidate review committee. This committee looks over the packets for each candidate, raises questions and concerns, and ultimately makes the final decision of which candidates to hire and which to reject.
The idea is that multiple interviewers each giving their own opinion of a candidate greatly reduces the chance that any individual interviewer’s bias will affect the outcomes for any individual candidate. If one person thinks a person can do a job, their judgment might be skewed in some way. But if eight people think a person can do a job, there’s a pretty good chance that they can actually do it.
In an ideal world, Facebook interviewers would be test-administering robots that can ingest the data from the interview and then compare that data against the data for every other candidate. Facebook would then hire the top ten percent of candidates.
This system works in practice but has one huge potential flaw: we can’t know whether it actually works in the long run.
The Problem with Facebook’s Hiring Process
A hiring model is only good if companies can refine it and improve the model over time. This means measuring the efficacy of their hires and correlating them back to their interview performance.
But each company can only tune their hiring process so much. While companies can validate if their interviews are a good signal by looking at true positives and (good hires) and false positives (bad hires), this also requires them to be able to validate their hires as high or low performers in a non-biased way as well.
Here, we can use a graph to illustrate the point:
Let’s say that Facebook looks at this graph and correlates it with their interview process and finds that the “good” hires all scored higher on the Fit metric than their “bad” hire counterparts. They might conclude that they should emphasize Fit more in their future hiring process.
But this only paints half of the picture because they’re only using true positives and false positives to generate their impressions.
To get the true negatives (bad candidates that got rejected) and false negatives (good candidates that got rejected), a company would have to somehow follow a person’s career after they failed the interview. Only then could they really understand if the person they rejected ended up being a “good” or a “bad” hire somewhere else.
It could be, in the long run, that there are plenty of candidates who scored poorly in the Fit category but wound up generating value at other companies.
In other words, the hiring process becomes a self-sustaining feedback loop that doesn’t take into account the bigger picture of what constitutes a “good” or a “bad” hire.
And so ultimately their interview tests and hiring models are only going to be mediocre (or worse) in the long run, given that they lack that extra information.
Don’t take rejection by a big tech company too personally. Because that rejection doesn’t define you or mean that you wouldn’t have been a good hire if they’d decided to make an offer. The important thing is to keep practicing the interview process, keep upskilling, keep doing the best you can at whatever it is you do.
Because, at the end of the day, you want to make it so that not hiring you might be the worst mistake a big tech company ever made.
Thanks for Reading
We hope that you found the article helpful. If you did, subscribe to the Interview Query blog to receive weekly updates and new articles to keep you up-to-date on the state of the data science profession and loaded with tips and tricks to keep you one step ahead of the competition.
If you're currently in the process of interviewing for a job in the data science field, or just looking to brush up on your DS skills, consider signing up for a free account on Interview Query for access to our question bank filled with real interview questions from companies like Google, Facebook, Amazon, and more. Our questions are sorted by question type, position, and difficulty to make sure that you're practicing exactly what you need to practice to level up your skills and ace your next interview.
Finally, if you're looking for an all-inclusive starter course for data science, look no further than the Data Science Course on Interview Query. Covering subjects ranging from Product Intuition to Machine Learning to SQL, our course will bring you up to date on the things you need to know to become a data scientist.
Take the next step in your data science career. Try Interview Query today.