Dispelling myths about interviews at FAANG

Dispelling myths about interviews at FAANG

Everyone has probably heard about the FAANG (and other BigTech) interview process. Leetcode tasks, system design, behavioral issues, culture fit, and other buzzwords are regularly discussed on Habre. The collective mind of Habra has more or less come to a general consensus: this is a ridiculous procedure designed by incompetent people.

As fate would have it, in my ten years of experience in FAANG+, I have been fortunate enough to discuss the topic of recruitment with many high-level engineers, managers, directors, as well as members of recruitment decision-making committees from several FAANG companies. With this article, I want to try to convey their position to a wider audience.

A widespread and approved opinion.

Recently, an article was published on habra, which calls for the eradication of the so-called Leetcode-interviews. For some reason, FAANG engineering hiring decision-makers are slow to accept this offer. Why do they need Leetcode interviews? There are many myths on this topic. My favorite versions:

  • Stupid Olympians hire stupid Olympians

  • It is necessary to overwhelm the candidate in order to offer him a lower salary

  • Those companies simply do not know how to hire developers and therefore offer meaningless tasks for “code that is already in the standard library”

And even more adequate options, such as the fact that everyone at Google writes a bunch of algorithms twice a week, are also far from reality: even if they really do write, this is not the point of the Leetcode interview.

Most people don’t understand how FAANG recruits. Yes, we all see what Google does, but we don’t know why it does it. An outside observer sees the section’s Leetcode and believes that FAANG employees are constantly faced with algorithmic tasks. Another observer knows that this is not the case, so he assumes that the interview is designed by incompetent people, because it does not test job skills. The third thinks more deeply and concludes that solving artificial (and useless in practice) tasks is a kind of stress test that tests the candidate for obedience and loyalty. Other people, often even getting a job at FAANG, but never understanding why they were hired, say that it’s just a “not-so-secret handshake” that you just have to learn to demonstrate your desire to get a job there.

All these thoughts are regularly flashed in articles and comments on Khabra. But what do the people who designed and have been supporting and improving this process for decades think about Leetcode and other sections? Let’s understand each other.

Myth 1: The recruitment process is made “as hard as it gets”, the FAANG has so much money that they don’t care how to hire.

The hiring process is definitely not “how it will work out”. Highly experienced directors and engineers with thousands of years of combined experience sat down together, clearly defined hiring challenges, and carefully designed a process that would address those challenges. And then they regularly evaluate and adjust this process for more than two decades.

Myth 2: The recruitment process at FAANG is not working well. Here I had a colleague at Google, and he was a bad developer.

First, the recruitment process at FAANGs is highly scrutinized, and works very well – for FAANGs.

Second, your colleague at Google can really be a bad developer and still be a good fit for Google. We will talk more about this below.

And thirdly, of course, sometimes this process makes mistakes and these mistakes are often very visible. But the presence of roughness does not mean that there are no clear goals or there is no clear and working plan to achieve them. And yes, the budget for mistakes is included in this plan.

Myth 3: FAANG has so many candidates that the interview process has to be made worse for scaling and lawyers.

There was a colleague’s article on this topic recently. Don’t get me wrong, the article describes everything correctly. All these details do have a place and require solutions, and the current process does more or less solve them. But this is not the essence of the problem. If the current hiring process did not achieve the set goals, it would be immediately changed to one that did, and the new process would be made as objective as possible and scaled up.

So, to understand why the hiring process looks the way it does, you must first understand the goals of its authors. What are the goals?

The easiest answer to this question is the obvious answer: the goal of the FAANG hiring process is to hire exactly the people the process hires. In other words, see who’s passing by, pick out a shared profile, and you’ll know who they’re looking for. In order not to dwell on such a long answer, let’s go through several clear signs.

Myth 4: FAANG hires Olympians who don’t understand anything about the real world.

The successful candidate is required to quickly understand incomprehensible, confusing situations. Leetcode tasks (usually easy and medium, there are simple and hard ones) do an excellent job of checking this point. You are given an intricate puzzle to “unravel”. But the purpose of the task is not to check the knowledge of algorithms. Algorithms are a nice bonus, the icing on the cake. The goal is to find a person who quickly understands incomprehensible situations — a very important skill for the real world. System Design interviews test this quality just as well, as the candidate needs to quickly understand some business requirements.

Myth 5: FAANG hires developers.

Yes, you did not hear. FAANG is not looking for developers. Many believe that if the position is called “Software Engineer”, then they are looking for a programmer. This is simply not the case.

The candidate is required be able to program — communicate with the computer in its language. They do not need knowledge of specific languages, libraries, frameworks, etc. It is not necessary to be a professional developer. Leetcode tasks are a great indicator of whether a candidate can explain to a computer what the candidate wants from this computer, and the above-proven skill of quickly figuring out an unclear situation allows you to learn any specific API, from a library to a programming language and to the heat of the railway

Myth 6: The giant FAANG machine hires small cogs.

On the contrary, the cogs are just weeded out at the interview. For successful passage, a candidate needs a quality that in FAANG is called “Leadership“. Everyone has probably heard of the Amazon Leadership Principles. In fact, the entire FAANG+ values ​​these qualities, Amazon is just the loudest about it.

The Russian translation of this word as “leadership” does not seem to me, as a native speaker of the Russian language, to reflect the essence of the matter. In Russian, “leadership” means “leadership over someone.” In the Russian language, they cannot be leaders at all. In the FAANG dictionary, this word refers to many things. To my taste, this word is easiest to reveal as “making shit happen” the ability to do things, overcoming difficulties.

Sometimes the difficulty lies in the need to organize people to solve complex tasks, but this is a very advanced level. The first difficulty that an engineer of any level faces is the banal ignorance of how to solve a specific task. Roll up your sleeves and figure things out – this is also called “Leadership”. As you might have guessed, Leetcode tasks are great for testing a basic version of this quality, demonstrating how a candidate would behave when faced with an unfamiliar problem. But the main test of leadership qualities, of course, is the interview of behavior.

At this point, many would think that the Leetcode interview preparation industry has zeroed in on this idea as the tasks are now familiar to everyone. Don’t jump to conclusions so quickly. The candidate can solve the task on the litcode as many times as he likes, and even learn it by heart, so he will not understand what the hiring committee expects from him.

Myth 7: FAANG is looking for the best engineers.

FAANG obviously wants to hire the best possible engineers, that’s true. But this is the main goal. First of all, FAANG is looking for people who will be comfortable with FAANG. Neither the company as a whole, nor specific future colleagues want a fresh person who has joined to dislike their job. No coolness can compensate for genuine interest. This is the essence of the so-called “Culture fit” interviews, but in fact, whether a candidate is a cultural fit is generally checked by all sections.

For example, everyone knows that Google is constantly closing projects. Not only large products like Google+, but also to the smallest project within the company. This is not accidental. Closing projects is a publicly declared value of the company: “We open easily – we close easily.” Everyone decides for himself whether he likes it or not. But regardless of whether you agree with Google, it chose this strategy and has been following it quite successfully for 25 years. Employees who don’t agree with this culture will spit on Google every day when their projects are closed. Google, on the other hand, publicly declaring its culture and values ​​at every corner, literally asks those who disagree not to come to interviews. But they still come and the company needs to filter them.

Myth 8: You have to grit your teeth and go through this ridiculous FAANG recruitment pipeline to get your dream job. Hard in training is easy in battle!

This is not a problem at all. The entire interview process culture fit check. If you hate the hiring process at a FAANG company, you won’t like working there either.

Myth 9: Google is successful, so it hires well, you should hire the same!

Understanding what qualities FAANG companies are looking for in candidates, it is clear that all interview sections are specially selected to test these very qualities. It is also not surprising that many developers find this process useless. Of course, it is useless if the task is to hire developers. If your company needs developers, please don’t copy Google. It won’t work.

In fact, it is not myth 9: Google is successful, which means it hires well, it is necessary to hire in the same way!

Copy Google, but not in the way you think – set clear goals for your company and design the hiring process according to those goals, like Google and other BigTech companies do. A thoughtful approach to hiring is a big part of success.

I hope this article helps demystify the FAANG recruitment process and help you understand if you even want to get there, and if so, what qualities you need to show in interviews. Whether you like the system or not, it’s useful to at least understand it. And for hiring managers, this article will help you describe who such a process is designed to hire and whether you need it.

I reveal this and other topics related to recruitment and career in FAANG companies from the point of view of decision-makers in more detail in my Telegram channel. In recent posts, I understand in more detail the responsibilities of a Software Engineer in FAANG, and also explain why Leetcode sections are also behavioral.

Related posts