Companies and interview loops

Company culture and questions to ask in a PM interview

Lesson 26 of 45~11 min read
In this lesson6 sections

An unofficial lesson drawing on the methods in the Company Research chapter of Cracking the PM Interview by Gayle Laakmann McDowell and Jackie Bavaro (2014). PracHub is not affiliated with the authors or publisher. Buy the book

Near the end of almost every PM interview you'll be asked what questions you have. The one you pick shows how well you read the company, and the answer tells you whether you want to work there. Both depend on understanding the culture before the loop starts.

Most candidates write their questions last, on the morning of the interview. This lesson writes them first and works back to the evidence each one needs. We'll follow Hana, the invented candidate from the research lesson, whose PM loop is with Spotify's audiobooks team. She has four years as a PM at a mid-sized fitness app, and her one-page Spotify brief is done.

Start with the questions you'll ask each interviewer

Expect a slot for your questions in nearly every interview, sometimes 10 or 15 minutes. Running dry after one leaves a silence the interviewer may recall when writing feedback. So Hana writes three questions per interviewer first. Each should get her something she needs and show she reads Spotify well.

To match questions to people, she looks up each interviewer's professional record: past roles, talks, posts about their work. That shows which questions suit whom, though not what working beside them is like. Anything personal she finds stays unmentioned, since raising it would feel like being watched.

InterviewerQuestionEvidence it needs
PMWhere did the last big project on your team start: a team proposal or a company bet?Where ideas start
PMHer guess about the audiobook hours allowance, worked out belowA business decision from her brief
Engineering lead, then designerThe same first questionA second and third view
Engineering leadMake It Happen is a stated value. When speed and reliability pull apart, who decides?The published values
Engineering leadHow much do PMs here prototype with AI tools, and does it help you or create cleanup?A line in her posting
DesignerAudiobooks share Home with music and podcasts. When this team wants more room there, who settles it, and against which goal?A culture trait and the business fact behind it
DesignerHuman Judgment is a stated value. When a test result and a designer's judgment disagree, which usually wins?The published values
Hiring manager or more seniorAudiobooks are going from 22 markets to over 180. My guess is this team now works more on markets with few local-language books than on new features. How close is that?A company bet from her brief

All three get the same first question on purpose. If each describes a team proposal that won leadership over, work starts with teams here, and Hana should bring ideas of her own. If each names a company bet handed down, her ideas should serve that bet. If the answers split by seat, that's news no single answer could give, and she'll ask the hiring manager which account is closer. How PM roles differ by company reads the same signal from a job posting. In the loop she also listens for who has the last word on scope, and how PM and design settle disputes. What happens in a PM interview loop covers the rounds.

The engineering lead's prototyping row comes from her posting. "Product manager" covers very different jobs in 2026, from AI product PMs who judge model quality to technical PMs who work on APIs. Say Hana's posting asks for experimentation and SQL and never mentions code. Her reading: she'll pull her own numbers and design tests but won't write production code, so her stories should show her querying data herself. Whether PMs prototype with AI tools here is open, so she asks.

The hiring-manager row looks ahead: senior people spend their week on where the company is heading, and a guess about one named bet gets them talking.

If her planned questions get answered early, she has spares ready:

  • When someone here was promoted recently, what had they done?

  • Who outside this team can delay a launch?

  • Which problem does this team keep coming back to without solving?

  • How has your own week changed since the team started using AI tools?

  • Which decisions still wait for a day when everyone is in the same room?

Turn a generic question into a hypothesis

The strongest questions carry a guess about a business decision. A guess shows you understand the market and gives the interviewer something to agree or argue with.

Hana's first draft of her second PM question was:

"How do audiobooks fit into your business?"

Anyone could ask that, and her brief answers most of it. Her rewrite starts from a choice Spotify made:

"Audiobook hours now come with the main Premium plans, unused hours lapse each month, and heavy listeners can buy top-ups. My guess is that the allowance is there mainly to make the whole plan harder to cancel, and the monthly cap keeps what you pay publishers predictable. Is that close to how your team is measured, or does the allowance have to pay for itself?"

It shows its evidence and risks a guess, but leaves room to be corrected. Even a wrong guess pays: if the allowance is judged on its own margin, Hana learns how her ideas will be measured.

Another draft asked "Why don't you have...?" about a missing feature, which grades the roadmap. She rewrote it as "What did it lose out to?", which assumes a trade-off and asks about it.

The night before, Hana reads her questions aloud and cuts any that a search would answer.

Explain one culture trait with the business model

A culture trait usually has a business reason that predicts how the company decides. Pick one trait, find the fact behind it, and say what it predicts.

Hana picks One Team, the value about shared goals across teams. The fact behind it is in her brief: music, podcasts and audiobooks are sold as one plan and compete for one Home screen. A listener who stays for audiobooks still pays for music, and a Home slot given to a book is taken from a song.

Her hypothesis, to test rather than state: when every format shares one bill and one screen, no format team can win alone, so goals are probably set on the whole plan, such as subscribers who keep paying, more than on each format's hours. If so, an idea that takes Home space from music needs the music team's agreement, and one that keeps whole-plan subscribers longer has the easiest path. Her second designer question tests this.

Mission-led companies need the same reading. When the mission leads the careers page, expect questions on it, and use the business model to see which parts the company pays for.

To practise a mission question, try Airbnb Mission, Belonging & Hosting.

Bring one change, sized for this team

Come in with one thing you'd change, starting from what users complain about most. The one-star reviews in Hana's brief had no audiobook group, so she adds a pass over audiobook reviews. Say the commonest complaint there is running out of hours mid-book.

First she checks what exists. Listeners can buy top-up hours, the app tells them when they run out mid-chapter, and the help page sends anyone checking their remaining hours into account settings. Her idea picks up where those stop, framed as a hypothesis, since the interviewer may own this area and know limits she can't see:

"You already sell top-ups and tell people when they run out. What I'd want to test, and you may know why it won't work: show the hours left on a book's page before someone presses play, and flag it when the book runs longer. I'd judge it by books finished, and watch that top-up sales don't drop. What am I missing?"

Then she sizes it. Spotify makes big bets, such as podcasts and then audiobooks in Premium, and runs small tests all the time. Its leaders are evidence too: since January 2026 its co-founder and former CEO has been executive chairman, with co-CEOs from product and technology and from the business side. Hana's reading: product keeps a voice at the top, where big bets probably start. At a startup, look at the investors too, and what else they've backed.

So she prepares two versions: the hours warning tested in one market, and a larger one letting unused hours carry over a month. Her third PM question asks which size the team funds.

Find out what an ordinary week there is like

This research is for Hana's own decision. Her brief is full of prices and margins and silent on the week she'd have there. The careers page exists to attract her, so she matches each thing she wants to know with a more candid source:

What she wants to knowWhere she looks
Which values are currentThe company's own culture page, read on the day she preps
What it rewards in practiceFounder interviews, and how it builds and ships products
How often plans and teams changeGlassdoor, Blind and LinkedIn posts by people who left, read for patterns, not one angry post
What the loop testsReported questions, such as PracHub's Spotify product manager questions
Hours and launch pressureA call with someone she knows who works there now
Parental leave, and how Work From Anywhere applies to this teamHer recruiter call
How soon she could move to another teamThe hiring manager, once she holds an offer

The first row nearly caught her out. Several guides she found listed five Spotify values, while the company's culture page now lists three: One Team, Human Judgment and Make It Happen. A story built around a retired value would have shown her research was secondhand.

The second row finds values no page states. Spotify ran audiobooks in 22 markets before expanding to more than 180, so it likes to roll out in stages and read the numbers before going wider.

Say the pattern in the third row is a reorganisation every year or so, with roadmaps reset each time. That weighs more with Hana than any perk: at the fitness app she owned one area for four years and saw her bets through. It doesn't rule Spotify out, but it adds a hiring-manager question:

"When this team's goals last changed, what happened to the work already under way?"

The last two rows stay out of the loop. Asking about benefits, pay and remote work is normal now, and the recruiter is the one to ask. Before an offer, a team-move question could read as one foot out the door. For more questions that only help you weigh an offer, see questions that help you assess a role.

To practise reading values and a mission, try Netflix culture fit: name a culture memo principle that resonates and back it with a story, Twilio culture fit: the company value you identify with most and the one that would stretch you or Connecting OpenAI's Mission to Your Own Experience and to Product Decisions. Read the company's current page before you answer.

Map each value to a story you can tell

Interviewers at many companies score you against the published values, and at some the values settle arguments in meetings too. Hana tags each reported Spotify behavioral question with the value it seems to test, so she knows which values need a second story. Every value gets at least one:

ValueA question that might target itHer story
One Team"Tell me about a time you needed another team to ship."She lent an engineer to payments for two sprints so a billing fix shipped first. Her feature slipped a month; failed renewals fell from 6% to 4%.
Human Judgment"Tell me about a call you made against the data."She stopped a winning sign-up test because complaints about surprise charges doubled, and owned the call in the review.
Make It Happen"Tell me about shipping under pressure."She cut a workout-sharing feature to one screen and shipped in three weeks. In month one, 9% of weekly users shared a workout.

Knowing the culture changes how she tells each story. With One Team as a value, her answer on influencing peers or rolling out an unpopular decision should show her sharing context and giving something up. In the loop she listens for the value a question targets and names it: "This is the decision I owned against the test result."

Product answers can name a value or published design principle too. Asked to improve audiobook discovery on Home:

"I'd add a shelf picked by editors beside the ranked rows. A new book has no listening history for a model to learn from, and with Human Judgment as one of your values, a person vouching for it fits how you already work."

The reorganisation pattern changes her map too. Teams that reshuffle often, like fast-growing companies, lean harder on behavioral rounds, so she adds two stories: work nobody owned that she picked up, and a plan dropped mid-quarter.

A company's origins shape what it screens for: builders where engineers founded it, taste where design leads, customer stories where the customer comes first. Take the current signal from the values page, not an old reputation.

Fit is also judged outside the questions: how you treat the coordinator, or react when an interviewer runs late, reaches the debrief.

Amazon is the clearest principles-driven case: 16 Leadership Principles, with interviewers usually assigned a few each. Use Hana's map at a larger scale; Amazon Leadership Principles and STAR stories covers the list. At a customer-first company, when a story's trade-off is unclear, side with the customer outcome over a quick business gain.

To practise, answer Amazon PM Behavioral: The Leadership Principle That Resonates Most With You, and Why and Amazon PM behavioral: discussing your weakest-fit Leadership Principle honestly. Then build a map for Amazon Leadership Principles Behavioral Scenarios before writing any story in full.

All career and values questions