Interview with Thomas Brouwer, Yelp Product Manager - 2

Click for the first part of the interview!

We had a very informative and enjoyable chat with Thomas Brouwer. He is working as Product Manager in Yelp. He is responsible for User Location Intelligence Team. If you are interested in being a PM or just want to wonder more in the world of PM, you should definitely check it! It’s also very informative about Oxbridge, PhD in UK, internships and many more!

Second part will be focusing Phd life in Cambridge and Product Manager

6 — After finishing your bachelor, you continued at Cambridge as a Phd student, what was your motivation? Did you think to start working or make masters? How was the application process?

In the UK, lots of people start working straight after their 3-year bachelor’s. I didn’t feel ready to start working yet, and my parents also advised me not to — in the Netherlands, everyone does a 3 year bachelor’s and 2 year master’s, and then starts working. So I decided to do a further degree.

When I was looking at options, I found out it was sometimes possible to go straight into a PhD program. I knew software engineering wasn’t really my thing, but maybe research would be, and I had done some machine learning in my undergraduate and liked it. So I started talking to professors in my department about this possibility. Because my grades were good, the two professors I talked to were both willing to support my application for the PhD program, so I decided to apply for a PhD in bioinformatics and machine learning, in the computer science department. This involved writing a research statement — just a couple pages stating what you are planning on researching in the PhD. The joke I always make it that on the first day of the PhD, you print that statement out, tear it up, and throw it away. It’s not going to dictate what you work on, it’s just an exercise to prove you can think like a researcher. My supervisor helped me a lot with this.

In the UK, some PhD programs offer funding directly, so getting into the program means you also have funding. However, many PhDs (including mine) do not include funding, and then either your supervisor needs to have it for you (mine didn’t), or you need to apply for funding yourself. My department had a few fully-funded scholarships from the UK government, given out to current Cambridge students based on their grades. So I knew that if I wanted to get the funding, I needed to do well in the exams.

At the same time, I also looked at master’s programs. I applied to one at ETH Zurich, and got accepted, but did not secure a scholarship to study there, so decided against it (Switzerland is an expensive place to live). So all my hopes rested on the funding. Fortunately, all went well in the exams and I was able to start my PhD.

7 — How was your PhD years? Did you enjoy research and thought about continuing research? What did PhD teach you?

My PhD years were amazing socially. Because I stayed at Cambridge, I was still part of a college (Homerton), and joined the graduate student body for organising events and welcoming the new students. I met many amazing people there, and developed a lot of social and organisational skills there. That’s the nice thing about a PhD — it gives you a few more years of being a student (at least in the UK; in some other countries it’s more like a job) and hence lots of time to take on new hobbies, make new friends, etc.

Academically, it was harder. I realised fairly early on in my PhD that research wouldn’t be my thing either. The things I did not like about software engineering — not thinking so much about why you’re solving a problem, but just solving a problem given to you, and then spending lots of time debugging and things not working — came back in research. My supervisor was also very hands-off, which is great because it teaches you to be an independent researcher. But if you’re new to research, the learning curve can be very tough. I ended up reaching out to a postdoc in the engineering department, and collaborating with him on my first paper, and he taught me a lot. Later in my PhD, I also visited a research group in Finland, and found out that others did actually think my research was relevant. So this experience taught me to be very assertive, self-driven, and to deal with uncertainty and stress.

All in all, I am very happy I did the PhD, because it taught me a lot about motivating yourself and being structured when no structure is given. Would I do it again? Probably not.

8 — You made two internships at Yelp as a Data Mining Intern and a Product Management Intern. How was Yelp? Why did you decide to make PM internship and how was the process? What were your takeaways from your internships?

As I said before, I realised early on in my PhD that research would also not really be my thing. However, I did enjoy the machine learning field, so perhaps doing machine learning in industry would be a good option. And how do you test that assumption? By doing an internship of course! So during the second year of my PhD I started applying to internships for the summer. Yelp had a stand at my department’s careers fair, and I thought “why not” and applied. I also had a friend who had previously done an internship there and really enjoyed it, so I knew it would be a good place to work.

Immediately the next day after the careers fair they did a first-round interview in Cambridge, and after that I had two more interviews over the phone, after which they offered me an internship position. That summer I paused my PhD for 3 months and spent that time working for Yelp in San Francisco, as a data mining engineer. This position means you write code like a software engineer, but you also do data analysis or machine learning. The team I ended up in was fairly new, so there wasn’t much machine learning yet, but I did get a taste for what it would be like to work in industry in a data-focused role. And again, it didn’t really click with me.

One of my good friends during my internship was a product manager, and we started talking about the role. I had heard about it before — another friend during my undergrad did some PM internships — and it had always sounded interested, but I never tried it. So while in San Francisco, I started talking to different product managers at Yelp to better understand the position.

I realised that this role would fit me a lot better: you spend most of your time asking “why are we building this feature”, rather than “how are we building it”. I reached out to the recruiter at Yelp responsible for these roles, and after a quick chat she agreed to let me interview for a PM position. She did warn me that once you switch to being a PM, you will probably never write code again. I was fine with this. Because I did not have any work experience, and I had time to do another internship after my PhD, I asked to be interviewed for an internship, rather than a full-time position. This probably set the bar much lower for them to give me a chance to experience what this role is, and also let me try it out without immediately taking a full-time job.

To prepare for the internship, I did some mock interviews with friends, and read the “Cracking the PM interview” book. The most important thing to do is to get into the mindset of a PM. When you get a question like “We are thinking of adding feature X to this page, how would you do this”, an engineer might say “Oh cool there is this new technology we can use to build X” and dive into the implementation details. Instead, a PM will ask: who is using this feature? What need or pain point would it solve? How would I measure whether this improves anything? How can we test this as quickly as possible? If that’s the kind of thinking you like, product management might be for you!

9 — You finished your PhD and started as a product manager at Yelp, that’s great! What are your tasks as a PM? Are all PMs the same, or do they focus on different problems?

Product management is very different from software engineering. Your job is to think about the consumer, and the problems they have. You will spend almost all of your week in meetings, interviewing customers, or thinking about projects and ideas, rather than writing code. Sometimes you might write some SQL queries to back up your ideas with data, but generally you will be using your social and project management skills more than anything.

Typically, each team of 6–10 software engineers has one engineering manager, who is responsible for the efficiency and career progression of the engineers, and one product manager. As the PM, it is your job to figure out what the team should work on. What should we do in the next month? What are we doing next quarter? What’s our vision for the next one to two years? Where are our biggest opportunities, what are our biggest challenges?

You will spend part of your time figuring out what projects you could be doing, discussing these ideas with other PMs, sizing the opportunity (e.g. “Building feature X can drive metric Y by Z%”), and writing out product specifications (or “specs”). You will also spend a lot of time project managing what the engineers are working on, making sure nothing is blocking the team from shipping the project as quickly as possible, and solving issues that the customers and clients might have with the product.

Effectively, a product manager operates at the intersection of technology, user experience, and business. Each PM might focus more on one of these aspects than the others, which I like to classify as follows. First of all there is the User Experience PM. They focus mainly on consumer-facing features, working extensively with designers (or at start-ups, sometimes even as the designer) to build the right features for the users. The second type is the business PM. They tend to come from business backgrounds, such as consulting or having done an MBA. They tend to focus on teams and questions regarding marketplaces and pricing. Finally, there are the technical or data PMs, who focus on more data related problems (such as machine learning), and work a lot with data scientists — or analyse data themselves — to identify issues and opportunities.

During my internship, and now full-time position at Yelp, I have touched on all three of these, although I am closest to a data PM. This makes sense, given my machine learning background, but you still have to be careful not to pigeon hole yourself into a specific mindset. For example, most recently I focused a lot on the business side of my product, because that is where the biggest need was — I interviewed a dozen of our customers, I estimated the revenue impact of our product, and identified the biggest opportunities to improve how we deliver our product to the customer. Now I am focusing more on the data side — how should we improve our machine learning model? Where can our data be improved?

In any situation, a mindset of being “scrappy” is very important: you should try to test and iterate as quickly as possible. Doing a perfect analysis is nice, but if you can reach the same conclusion in 10% of the time with an imperfect analysis, you should definitely do the imperfect one. You can always decide to refine it later, but by then odds are, you are focusing on something entirely different. This can be hard sometimes, especially if you are a perfectionist like myself, but you will need to learn to do this if you want to be an efficient PM.

10 — As a final question, What do you suggest people who want to become Pm one day? What are the possible ways to become PM?

First of all, you should ask yourself whether you want to be a product manager. Are you okay never writing software again? How do you feel about spending a significant part (sometimes more than half) of your week in meetings? Do you want to do everything perfectly, or do you want to be scrappy? How do you deal with failure and uncertainty?

The best way to test this is doing an internship. But you can also grab a coffee with product managers in your environment and ask them about their experiences, what they like and dislike about their jobs, and see if it resonates with you.

Let’s say you’ve decided you want to become a product manager. It’s not always easy to kickstart a career in product management at a tech company. Not many large companies will hire inexperienced new graduates straight out of university. Typically, companies will be looking for a few years of work experience in a consumer-facing position, such as customer support, or business analytics. They want you to demonstrate that you can think like the customer, so empathy is very important.

In my eyes, these are the best paths into a product management position:

• Do an internship or two as a product manager, and try to convert that into a full-time job offer at the same company.
• Join a start-up as a product manager. They are more likely to take on someone inexperienced. The downside is that you will not get much guidance and structure on how to be a good product manager, which can be a problem (especially if you are inexperienced).
• Work as a software engineer for a few years, and then try to convert to a product management position at the same company — if you have proven your worth as a good employee, they are more willing to give you your shot. Try to get more project lead positions, and ask to PM a small project. This will let you try the role, and prove that you can do it.
• Work in business analytics or customer support for a few years, and then apply for a PM role.

As the saying goes, “all roads lead to Rome”. There are many ways to get into product management. It might not always be possible to go straight into PMing out of university, but there are many ways to get into it after a few years of work experience, so don’t despair. One of those roads will lead you into product management.