The round types and how they differ

#Why the round type matters

Most PM interview loops are made of several rounds, each run by a different interviewer and each built to test a different thing. The single most expensive mistake you can make is answering the wrong kind of question: treating a metrics question like a design exercise, or turning a strategy question into a feature brainstorm.

The good news is that round types are easy to spot. The first sentence of the prompt almost always tells you which one you're in. This lesson gives you a field guide.

A caveat before we start. Companies name, combine and order these rounds differently, and they change their loops over time. Some fold estimation into a metrics round. Some run a single "product" interview that covers design and metrics. Smaller companies may replace several rounds with a take-home exercise or a presentation. Treat the categories below as the building blocks most loops are made from, not as any one company's process. How interview loops differ covers the variation.

#The field guide

#Product sense and design

Sounds like: "Design a product to help people plan group trips." "How would you improve a public transit app?" "Tell me about a product you love and how you'd make it better."

This is the round most people picture when they think "PM interview". You're given an open prompt and asked to design something out loud. The interviewer wants to see you set a goal, pick a specific user, find a real problem, generate a few ideas, choose one with a clear reason and say how you'd know it worked.

What separates strong answers: a narrow, well-justified user choice and a pain point that feels true. The solution matters less than the path to it.

Where to learn it: Product sense and design, module 1.

#Metrics and execution

Sounds like: "What's the north-star metric for a podcast app?" "Daily active users fell 10% last Tuesday. What do you do?" "This A/B test raised clicks but lowered time spent. Do we ship?"

This family covers three related questions. Choosing metrics: what should we measure and why? Diagnosing: something changed, find out why. Deciding: given this data and these trade-offs, what's the call? Some companies call this an execution round because it tests how you'd run a product day to day.

What separates strong answers: a calm, ordered diagnosis (is the data right? is it us or the world? which segment?) and a willingness to make a decision with incomplete data.

Where to learn it: Metrics and execution, module 2.

#Estimation

Sounds like: "How many electric scooters are there in Chicago?" "Estimate the yearly revenue of a US dog-walking app." "How much storage does a photo app need per year?"

Sometimes a standalone round, often a ten-minute segment inside another one. Nobody expects the real number. They want a clear breakdown, sensible round numbers, clean arithmetic and a sanity check at the end.

What separates strong answers: announcing the structure before doing any math, and checking the result against something you already know.

Where to learn it: Estimation, module 3.

#Product strategy

Sounds like: "Should a ride-hailing company launch grocery delivery?" "A competitor just cut prices by 30%. How do you respond?" "Should we build our own payments system or partner?"

Strategy rounds zoom out from a feature to a business. They test whether you can think about markets, competition, a company's strengths and how money is made. They are more common for senior roles, but many loops include at least one strategy-flavored question.

What separates strong answers: a clear recommendation, early, backed by two or three reasons and an honest risk.

Where to learn it: Product strategy, module 5.

#Behavioral and leadership

Sounds like: "Tell me about a time you disagreed with an engineer." "Describe a product that failed." "Why do you want to be a PM?" "Why this company?"

Nearly every loop has one, and many interviewers fold a behavioral question into other rounds too. You answer with a real story from your past. The interviewer is checking how you handle conflict, influence people you don't manage, own outcomes and learn from mistakes.

What separates strong answers: a tight story where your own actions are specific, the result is concrete and the reflection is honest.

Where to learn it: Behavioral and leadership, module 4.

#Technical

Sounds like: "Walk me through what happens when you type a web address and press enter." "How would you design the back end for a simple ride-booking app?" "An engineer says this feature will take six months. How do you respond?"

Not every PM loop has one, and the depth varies widely. For technical PM roles, platform roles and some large companies, expect it. You won't be asked to write production code in most PM loops, but you should be able to reason about systems, APIs, data and trade-offs, and show you can work with engineers as a credible partner.

What separates strong answers: clear explanations at the right level, and honesty about where your knowledge stops.

Where to learn it: Technical for PMs, module 6.

#AI product

Sounds like: "Design an AI assistant for customer support agents." "How would you measure whether our AI summaries are good?" "Our chatbot sometimes makes things up. What do you do?"

A newer round type, increasingly common for roles that touch AI features. It blends product sense with questions specific to systems built on machine learning and large language models: what the model can and can't do reliably, how you evaluate quality, what can go wrong and what it costs to run.

What separates strong answers: designing around the model's failure modes rather than assuming it works, and having a concrete plan for measuring quality.

Where to learn it: AI product interviews, module 7.

#Formats you'll meet

Round type is what you're asked. Format is how. Common formats include:

• Recruiter screen. A short call about your background, motivation and logistics. Covered in the recruiter screen.
• Phone or video screen. Usually one or two of the rounds above with a PM, often product sense or a behavioral question.
• The loop. Several back-to-back interviews, in person or on video, each covering one or two round types.
• Take-home or written exercise. A product spec, strategy memo or case you prepare in advance, sometimes presented live.

Try it Label each prompt with its round type before reading on. (a) "Our app's ratings dropped from 4.6 to 4.1 this month." (b) "How many weddings happen in the US each year?" (c) "Should a music streaming service start selling concert tickets?" (d) "Design a fitness app for people over 70."

Answers: (a) metrics, diagnosis. (b) estimation. (c) strategy. (d) product sense. If you hesitated on (a), notice that "dropped" is the tell: any prompt where one of your own metrics moved without explanation is a diagnosis question first.

#Hybrid prompts

Real interviewers mix types. "Design a feature to increase bookings for a hotel app, and tell me how you'd measure it" is product sense with a metrics close. "Should we launch in Brazil? Roughly how big is the opportunity?" is strategy with an estimation inside.

Handle hybrids by naming the parts out loud: "I'll answer this in two parts. First, whether we should enter, then a rough size of the market to back it up." That one sentence shows structure and stops you from drifting.

Key idea In the first ten seconds, name the round type to yourself and pick the method for it. The right method on an average idea beats the wrong method on a brilliant one.

#Recap

• Most loops are built from seven round types: product sense, metrics and execution, estimation, strategy, behavioral, technical and AI product.
• The first sentence of a prompt almost always tells you the type; an unexplained drop or spike in your own metric means diagnosis.
• Each round leans on different signals and sets a different trap.
• Companies combine, rename and reorder rounds, so prepare for the building blocks rather than one fixed loop.
• For hybrid prompts, name the parts out loud and answer them in order.