This article was written by Jevgeni Kabanov, President of Bolt. If you want to discover more of his articles, head to Jevgeni’s personal blog.
Standard interview structure
Over the years, I’ve developed a standard interview structure. It works best for senior hires. Junior hires should generally be interviewed for soft skills only. Otherwise, it’s pure testing through tasks.
The structure is simple:
- Questions to them
- Questions to me
I’ll follow roughly the same structure in this post. But first, let’s talk about the goals and methods of the interview.
Goals and methods
The interview’s goal is to decide whether a candidate is fit for a role (and if you have several positions available, then the secondary goal is to identify which role the candidate is a fit for).
Most of us have a model of an ideal candidate in our heads. I usually formalise that model, splitting it into components I can independently evaluate and score to make different candidates highly comparable. E.g. for product managers, I look at the following characteristics:
- Out-of-the-box thinking
- Culture fit
I can talk more about skill models in a different post. Still, I approach the hiring process and the interview to validate the aspects I identified in the skill model for the role.
I’ll take copious notes during interviews because I know the score won’t capture everything about the candidate. In two weeks, if I have to choose between several options, I’ll not remember the context anymore, and looking at my notes has helped me a lot in such cases.
Since the start of the pandemic, it’s become a lot easier because I haven’t had to explain my constant typing as much to the candidates!
Before the interview, we’ll always ask the candidate to complete homework. This usually takes 4–8 hours to complete and includes tasks similar to what the candidate would solve daily in the role.
There’s usually a bias against senior candidates completing the homework. Still, we’ve rarely compromised on this requirement — many candidates passed the interview process but failed the assignment.
At the start of the interview, I’ll introduce myself, ask the candidate to introduce themselves, and ask why they’re interested in the company and particular role. This is a way to kick off the conversation and discover why people search for a role. There isn’t a correct answer here, though you generally look for passion over pure mercantile interest.
The coolest motivation I’ve heard was from a determined guy who said, “I want to be a startup founder, and I’m convinced that joining your team is the best school I can get”. He had little experience and a different background but proved himself incredibly quickly.
I’ll ask the candidate to add colour to the CV I have in front of me. The goal is to provide the context for the rest of the interview. A CV and track record have almost no influence on my evaluation. I spend under 10 minutes on the first two parts.
I like CVs that are well structured around quantifiable deliverables, but often see ones disconnected from reality. If a PM has increased company revenue by 20% in half a year, there best be a solid story to defend it. Often, their activity was a minor part of an overall change in the business.
Here, I’ll ask a candidate to pick two projects from their past and go deep into the discussion. A project can be anything, with a beginning and an end, resources used, data gathered, decisions made, and the resulting impact. This is where the meat of the interview starts.
The goal is to get insight into the thinking process of the interviewee. I push them to get specific (I did X, then Y and Z, over I followed the ABC process) and try to think along and ask five why style questions to make them go deeper.
There are several aspects I can evaluate if someone isn’t right for the role:
- Storytelling: can they tell a compelling narrative? This is a surprisingly good proxy for emotional intelligence.
- Systematic: how structured are they? Is there a logical order to the story, and do they break problems down into parts and then arrive at solutions? Do they mention the fundamental drivers of their decisions? Can they explain them adequately?
- Creative intelligence: do they arrive at counterintuitive insights? What do they do with them? Do they iterate through learning and brainstorming to find new ways to achieve their goals?
- Collaborative: do they mention others? Do they attribute credit correctly or just focus on themselves?
Depending on the role, I’ll also listen for other proficiency signs in specific skills and ask open-ended follow-up questions. I try to avoid specific questions in interviews as they often give away the answer.
One thing I rarely test directly is people management, as I’ve found that people who are structured, smart, and have good emotional intelligence never have trouble managing others. Whereas a lot of so-called great people managers fail quickly as they don’t have enough skill in their existing domain. But that’s a topic for a follow-up.
You may wonder why I’m giving away my structure, making it easier for future candidates to pass my interviews, and that’s because of the next section.
I spend at least 10 minutes in an interview working on a task with the candidate. But in reality, that, combined with reviewing the homework, is often responsible for most of the results. It cross-validates everything I learned earlier, and I need it to have high confidence in the evaluated skills.
For the task, I pick some simple problem statements adjacent to the role in question (but not likely to have been encountered by the candidate previously). We have several tasks with the problem statements and evaluation guidelines that we rotate with candidates.
A (contrived) example for a product manager role could be “How would you sell airline tickets?”. This is a decent (albeit unverified) example of a simple question with a lot of depth. You should work out that you need to maximise profit per flight, which requires maximising utilisation, from which you can arrive at overbooking, dynamic pricing, and loyalty schemes, and even spare capacity auctions. Not to mention online/offline channel management and integration with unsophisticated third-party systems.
In the task, I’ll also prod the interviewee to get them to think deeper and describe the actual insights, not just the process. As in the Projects section above, I’ll examine how they attack problems and what skills they demonstrate in the process.
Questions to them
I’ll have a few open-ended questions prepared for some roles and sometimes pick up more of them during the Project and Task discussions. Sometimes I’ll also ask follow-up questions on the Homework.
Questions to me
I’ll always leave at least 10 minutes for the candidates to ask me questions. The interview is a two-way process; they should also learn about me. I’ll answer questions as openly and transparently as I can.
For example, someone once asked me if the reputation of Bolt as a copycat is true, and I answered, “Yes, and we’re proud of it”. I then explained that our only driver is to create value for customers, and if the best way is to copy someone, we’re happy to do so. But we’ll also differentiate where we think we can create more value.
And that’s it! At the end of the interview, I’ll score the different aspects and add a short write-up so that the hiring manager can make a good decision without further context.
This structure works for a surprising variety of roles as the projects vary (launching a country, spinning up a team), but the process remains the same.
We have hundreds of open roles, from entry-level to top-level management, each an exciting opportunity to contribute to making cities for people, not for cars.
If our mission sounds like something you want to be a part of, and if you’re ready to work in an exciting, dynamic, fast-paced industry and aren’t afraid of a challenge, we’re waiting for you!