Roberto Marin: “The company that really changed me was Amazon”

OK, great question. I think I was one of those kids who had lots of different dreams. I played guitar and had a band for 15 years. I am still playing, actually. Yesterday, we were rehearsing with my colleagues because we are going to play at an Uber party. There is a music event, Uberpalooza, which is like Lollapalooza but for Uber employees.

When I started university, I didn’t like computers that much. Networks, cables, hardware stuff, that part didn’t excite me at all. I always liked games and logic, so I really enjoyed building Lego and this kind of stuff, as well as puzzles and chess. So when I realized that programming is like building your own Lego with logic, I think I really liked it.

I think every company offered a different learning experience for me. But I think the company that really changed a lot of things, including helping me discover my own biases, was Amazon. It was my first American company, very self-service and with zero babysitting.

There, I also learned a lot about hiring. I think this is one of the processes that I admired the most at Amazon. It’s really serious. They are the best at removing bias that I have ever seen. Better than any other company.

Before Amazon, I always thought that I was great at hiring. And I discovered at Amazon that I was great at hiring brilliant people, but not hard workers. Brilliant people can be amazing. Like the Brazilian football team, they are always brilliant individuals. But building a team is different from having brilliant individuals.

Instead of asking during the interview what your perfect team would look like, they ask you about a specific team you worked with in the past and what you would have done differently.

The focus is always on things you’ve actually experienced and done. Every interview follows this real work experience approach: “Tell me about a time when something happened,” “Give me an example of an initiative where you faced this situation,” or “Tell me about the most challenging feedback you’ve received.” It’s never, “What would you do in these circumstances?” It’s always about what you actually did in a real situation.

And then you capture lots of things. If this is a junior or standard engineer, maybe they won’t have received strong feedback, and that’s fine. But as seniority grows, if that person says they have never received critical feedback, there are two possibilities. Either they didn’t recognise when they received an important piece of feedback, which is the case in most scenarios. Maybe they are closed off or arrogant, I don’t know; multiple things could be the case, so they don’t recognise that they are receiving important, critical feedback. Or somehow they are incredible performers, which is usually not the case.

The follow-up questions are very good. If someone is weak at describing their experience, they ask: “Why did you feel that this person was not doing a good job? Give me an example when the person didn’t demonstrate it. Didn’t they do the exercise? What happened?” And then it goes to the next level: “But this is a standard engineer. Why were you expecting them to deliver this result?”

And the hiring manager makes the final decision. They get all the evidence and, together with the bar raiser, reflect on it and come to the final conclusion. It’s not human resources or a team. Of course, the team and HR participate in the loop, but the hiring manager makes the final decision, and it’s on their shoulders. So they have to look at it because, in the end, if that person doesn’t work out, it’s their responsibility.

Also, during the evaluation, the vote is blinded. I don’t know HR’s or other colleagues’ scores before I submit my score. So this also removes bias because if you see everybody going, “Thumbs up, thumbs up, thumbs up,” then it influences you.

There are times when you have four huge thumbs up and one person giving a strong no. And then you would say, “OK, if it’s four thumbs up and only one strong no, of course the person is going to be hired.” No. Maybe the bar raiser will help the hiring manager understand that those gaps are significant for that particular role, and it’s not a fit.

Another thing Amazon doesn’t do is culture fit. Culture fit is a way for you to impose your prejudices on people and say they don’t belong there, whatever the reason. Amazon doesn’t have this. Or if they do, it’s in a way that’s almost impossible to remove completely because, in the end, we are human beings.

But when there is culture fit, I always feel like, holy smokes, really? It’s 2020-something, and we’re still calling it culture fit. Who doesn’t fit? If you are a burglar or a criminal, you don’t belong in any company, right? Nobody is going to hire you. This is not culture fit. This is: are you a good human being, a decent person in society?

I think the overall process at Netflix is less structured. It’s much more dependent on the hiring manager. The person who is going to be your boss defines the hiring process that suits them, so it’s very flexible.

But the accountability and responsibility at Netflix are even higher than at Amazon. If you make a wrong hire, it’s your call. And if you do it once or twice, you can be let go. So I would say it’s easier to hire someone at Netflix because you can tweak the process and do what works for you, but it’s also much easier to get fired if you make the wrong decisions.

A small example: at Amazon you get your badge, pick up a keyboard and go to your desk. If the next day you wanted a second keyboard, the system would say, “Hey, you already got a keyboard. Why do you need another one? Did you spill coffee on it?”

At Netflix, there’s no key card reader. You just grab one. But that freedom comes at a price. They don’t put many controls in place because they expect you to use good judgement.

I remember that even as an individual contributor, you could have a credit card and the authority to approve very large amounts. Someone on my team could technically book a business-class trip to the US and just go. But internally, we called that a “unique experience” because you’d probably only do it once. Then you’d be fired.

At Amazon, the leadership principles are very clearly documented. At Netflix, it’s more about cultural values, and one of the most important is good judgement. The idea is: would you make the same decision in your personal life?

For example, I was hired as a senior manager and even had director status for Netflix in Brazil, but I never took a business-class flight. I don’t fly business class in my personal life, so why would I book one just because Netflix was paying? I would rather fly economy two days earlier and spend the extra time in a hotel.

That’s the difference. At Netflix, you have access to everything. You book things yourself and make your own decisions. At Amazon, there were clearer rules. For example, you might have a $70 daily allowance. They didn’t care whether you spent all $70 on breakfast and nothing for the rest of the day, or spent nothing one day and $140 the next. There was flexibility, but there was still a defined limit. At Netflix, the rule was simply: use good judgement.

I think Netflix benefits from being a company of around 10,000 employees. They can keep processes very lean and put a lot of trust in employees.

But when you’re the size of Amazon, with around 1.6 million employees, doing that at the same level would be crazy. And even Netflix was beginning to realise that. As the company grew, they started introducing more controls.

When I joined Netflix, there weren’t even formal levels. You were basically an engineer, regardless of your level, and you worked things out with your manager, including your salary.

The salary philosophy was “personal top of market.” Essentially: “Natalia, can you get a better salary somewhere else? If you can show me that you can, I’ll match it immediately.”

It applied to employees as well. They even encouraged you to participate in external interviews and bring back offers from other companies. That was still the case during my first year at Netflix.

After that, things started to change. They hired Deborah Black from Amazon as VP of Engineering, and she introduced levels. Netflix had never really had that kind of structure before, and it was very unpopular. People didn’t like it because previously everyone was essentially on the same level. You didn’t necessarily know how much other people were earning or where they stood in a hierarchy.

Then they introduced levels and salary ranges and started thinking more systematically about career growth, promotions and other processes. The original model had worked for a long time, but as Netflix grew, they started changing it.

I still think Netflix’s culture is amazing. Being an individual contributor at that company is an incredible experience. But being a leader is difficult because things can be blurry. You’re expected to use “good judgement,” but different people can have very different ideas about what good judgement actually means.

That’s a good question, and I don’t know if there’s a simple answer. It’s actually very similar to your example of raising children.

I was raised with an incredible — maybe even extreme — amount of freedom, and it worked out. But did it work because I had so much freedom, or because of who my mother and I were? Probably the second. The same approach wouldn’t necessarily work for everyone.

I’ve also seen the opposite: people raised with a great deal of structure who turned out just as capable. Co-parenting has shown me that two very different approaches can meet in the middle and still work well.

So my answer is probably quite boring: for some initiatives, I would say Amazon has the better culture, and for others, Netflix does. Most people would probably prefer the Netflix style because it gives you so much freedom. But I don’t think that model works for everyone.

Exactly. At Netflix, there are very few guidelines around you. Of course, your manager will still give you feedback. They might say, “Other teams are struggling to interact with your team. What are you going to do about it?”

In that sense, your individual responsibility at Netflix is higher because it’s your responsibility to build that culture and trust. And if it doesn’t work, it’s much easier to get fired from Netflix than from Amazon.

At Amazon, if someone isn’t performing, there is a structured process. As a manager, you need to document the performance issues, talk to your HR business partner and build an improvement plan. Even if the person doesn’t succeed with that plan, the company may try to give them another opportunity in a different area. Maybe the problem is simply that this particular manager and employee don’t work well together, while that employee could be great somewhere else. Amazon knows how difficult it is to hire great talent, so there is some protection for the employee.

It’s an interesting contrast. Amazon can be much stricter when it comes to processes. If the rule says the limit is $70, then it’s $70. But when it comes to letting someone go, there is more protection for the employee. The manager has to provide a lot of documentation. That also helps reduce bias because otherwise a manager could simply decide they don’t like someone and fire them.

At Netflix, managers have much more authority. If you say, “I don’t want this person on my team,” you can work with HR to decide when they will leave and then have that conversation with them. But that freedom also comes with accountability. If you make that decision once or twice and it turns out you made the wrong calls, your own manager may end up firing you.

I think Brazilians are generally very flexible. We adapt easily, and we tend to like environments that give us flexibility, so in that sense, the Netflix model can be a good fit.

At the same time, I think we can sometimes get carried away when something exciting happens or when we suddenly become successful. Of course, this is a huge generalisation, and I don’t want to stereotype Brazilians, but I think there can be a tendency to get very excited about success and lose some focus.

You sometimes see it with young musicians, for example. Someone becomes rich and famous very quickly and suddenly gets carried away by it.

So I think the important thing is to enjoy that freedom and success while keeping things in perspective. You can think, “Yes, this is amazing,” but you still need to work, deliver results and take your responsibilities seriously.

When I was leading a team in France, I found it really challenging at first. People questioned every single decision down to the smallest detail. They wanted to understand everything, and sometimes I found that frustrating. I would think, “This is simple. Why do we need to keep discussing it?” We could spend three hours discussing one decision, with people questioning it again and again. Initially, I interpreted that as a lack of trust.

But then I noticed something interesting. After those three-hour discussions, I would hear people talking about the decision in the corridors, explaining it to others and supporting it. Once they were convinced, their commitment to the decision was very strong and sustainable.

My experience in Brazil was quite different. You could have a 20-minute conversation and everyone would be enthusiastic: “Great decision. Let’s do it.” But afterwards, when you weren’t there, people might say, “This is terrible. Why are we doing this?”

I’m exaggerating a little, of course, but the difference was noticeable. With the French teams I managed, it could take a long time to convince people, but once they were convinced, they would stick with the decision until it changed.

With Brazilian teams, I found that people were more likely to agree with you directly but express their doubts later. Because of that, I learned that I needed to overcommunicate. I would explain the same decision several times from different angles.

That way, if Natalia later complained about the decision to her colleague, they might say, “Yes, but remember this point Roberto made.” You create some connective tissue within the team that helps people understand why you’re moving in a particular direction.

I think it’s very human, and it’s connected to what I said earlier about flexibility.

If someone in Brazil says, “Let’s meet next week,” it doesn’t necessarily mean you have a plan to meet next week. It can simply mean, “I like you. It would be nice to see you again.”

In Rio de Janeiro, someone might even say, “Come to my house sometime,” but you don’t have their address or even their phone number. It’s probably not going to happen. But they’re not deliberately lying. It’s just a way of communicating.

When you grow up in an environment where things can be very flexible, I think you can develop a certain scepticism towards decisions. Someone announces something and your instinct is, “Yes, but it might change.” You don’t necessarily treat it as final.

I remember something similar happening with seatbelts in Brazil in the 1990s. They had already been mandatory for years, but enforcement wasn’t very strong. Then the authorities started taking it seriously, issuing fines and stopping cars, and people’s behaviour changed dramatically.

So mindsets and behaviours can change, but there need to be clear consequences. From my experience managing teams in Brazil, if you introduce a process, you also need to show that it actually matters and that there are consequences when people don’t follow it.

Exactly. Let’s say you tell everyone they need to update their OKRs by Friday. The first week, everybody does it. The second week, some people don’t. Then you need to follow up: “Where is this? Why hasn’t it been updated?” You either make the process automatic or consistently reinforce it until it becomes a habit. That takes time.

With the French teams I managed, my experience was different. Once people understood why we were doing something and agreed to it, they would keep doing it without me having to remind them.

In fact, they might eventually come back to me and say, “Why aren’t you following up on this anymore? You told us this was important, so why are we still doing it if you’re ignoring it?”

So accountability works in both directions, which I think is a good thing.

There’s one classic example. There was a very respected engineering training school here in Brazil. They were particularly good at finding and developing engineers. They would hire people straight from university, train them as interns and then convert them into engineers. For about 10 years, if you hired someone who had come through this company, you knew you were getting one of the best engineers around. There was one engineer from this company whom I wanted to hire. He had actually been fired from there.

I asked someone I trusted who knew his work very well for a reference. The answer was unusually blunt: this was one of the very few people they had ever let go.

And I thought, “Yes, but I’m different. I’ll manage this person differently, and he’ll perform.”

Why did I think that? Because he was a brilliant engineer. As I mentioned earlier, I was very good at identifying brilliant people and hiring them. But I was terrible at identifying hard workers.

It turned out that not only was he not a hard worker, but he also became a detractor of almost everything we were doing. He connected with another person on the team, and the situation became a complete mess.

Eventually, I had to fire three people on the same day. It really hurt my reputation.

This was at Farfetch, before I joined Amazon. Farfetch’s hiring processes at the time weren’t nearly as structured as Amazon’s or Netflix’s. I was the first hire in Brazil and was responsible for building the team. We grew to around 40 or 50 people, and I hired every single person on that team.

I did the best I could. We had interview panels, and colleagues from London and Portugal helped with the hiring process, but ultimately I was the decision maker.

So when those hires didn’t work out, the responsibility was mine. Having to fire three people at once was catastrophic for me. My reputation was seriously damaged for a while.

At the same time, I was in the process of being promoted to director. I did eventually get the promotion, but this situation undermined a lot of what I had been building. I felt terrible about it, and I’ve never forgotten that day. I still remember the exact date.

The lesson was: why on earth didn’t I trust the reference check?

I had actually done everything right. I did the reference check myself and received a very clear warning. But then I challenged the result of my own reference check. I decided to make an exception because I believed in the person.

I tend to be very human in the way I make decisions, and that can be a strength. It has helped me many times. But sometimes it can also put you in a very difficult position, and this was one of those cases.

Exactly, and it cost me a lot. It almost destroyed the entire team, and the damage wasn’t really fixable afterwards.

Even though I was promoted, things had changed. Then Amazon invited me to join their interview process. I never thought I would actually pass an Amazon interview, so I decided to take the chance.

I passed, they made me an offer, and I left Farfetch. If this situation hadn’t happened, perhaps I wouldn’t have accepted Amazon’s offer because my trajectory at Farfetch had been amazing until then. So in the end, maybe something good came out of it. But at the time, it was a terrible experience.

Natalia Domashenko: Thank you for your honesty. I think it’s really important to hear stories like this because sometimes you learn much more from failures than from achievements.■