Companies and interview loops
How to research a company before a PM interview
In this lesson10 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
Researching a company before a PM interview means building a one-page model of its business: who uses it, who pays, what serving them costs, who could take them away, and where it's heading. The page should end in one bet on what the company builds next, with the first two objections a skeptic would raise.
Our example is Hana, an invented candidate with a PM interview at Spotify next week. Every figure in her brief carries a date and a source, because Spotify changed its prices and restated its past results this year. The figures were checked in November 2026; recheck them before reuse.
Two product opinions you can defend
Interviewers often ask which of their products you like and what you'd change. Hana uses Spotify before reading about it, and by Friday her answers are specific enough to defend:
"My favorite feature is Discover Weekly: I've played it most Mondays this year and save about one song in four. I'd change the home screen. I come for music, and it often leads with podcasts and audiobooks. I see why, since time in any format makes me less likely to cancel, but I'd let a music-first listener choose what leads, then check whether they cancel less or just play fewer podcasts."
Discover Weekly comes from her Mondays on Premium. The home-screen complaint comes from one moment this week on a fresh free account with ads: she opened the app for a song and met podcasts first. On Premium she'd learned to scroll past them; as a newcomer she nearly closed the app. That moment outweighs a page of small gripes, because a real newcomer has no homework keeping them there.
Her fix would push podcasts down for music-first listeners, so she also opens a creator account and uploads a short test episode through Spotify for Creators, to see what podcasters stand to lose.
Her second opinion, on audiobooks, is smaller: she likes them sharing an app with her music, but would make the monthly hours limit easier to see.
The three accounts also show how the parts feed each other. Free listening brings people in, and some upgrade. One bill covers music, podcasts and audiobooks, so an hour on any of them makes cancelling harder. Every play trains the recommendations, which draw on what creators upload.
If the company sells to businesses through a sales team, you often can't sign up. Use what's public: documentation, recorded demos, review sites, and any customer you can talk to.
To practice this step, take Company Knowledge: Explaining Rakuten's Ecosystem, Business Segments and Strategy. Use two of its services if you can, note what pleased and annoyed you, then stop.
Plan for the Monday Discover Weekly breaks
Hana rehearses one failure: millions of Monday playlists come out stale, or packed with songs people already know. Her plan, in order:
Spot it within the hour: early skips and saves per playlist against last Monday's, plus social posts.
Protect listeners by serving last week's playlist. A stale list costs a little delight; a bad one teaches people to stop opening it.
Roll back whatever changed and rebuild the playlists.
Add a check that would have caught it before release.
A working bet: memberships for musicians
On the creator account Hana found memberships, which let podcast creators charge their most loyal fans. That gives her a bet before she reads a single filing:
"I'd extend memberships from podcasters to musicians. An artist could sell a monthly membership on their artist page, with members-only releases and listening sessions, and Spotify would take a share. A fan who pays an artist here has one more reason to stay, and that loyalty is to a person, which no AI-made track can earn."
The bet gives her reading a purpose: most sections below test it, and the answers go in the brief whether it survives or not.
Is the pain real?
Reviews show pain she didn't have to imagine. Hana sorts the 60 most recent one-star reviews in her phone's app store: 21 about ads and limits on the free plan, 14 about the home screen, 11 about playback bugs, 8 about price, 6 other. None of the big groups is about artists, so her bet doesn't answer a common listener complaint. Five-star reviews and threads on why people would never switch keep naming recommendations that find songs they love, and years of saved playlists. Her bet must not touch either. The subreddit and community forum show which complaints recur; for a business product, G2 and Capterra do that job.
The pain sits with two groups the reviews miss. One is superfans, the small share of listeners who would pay a particular artist far more than a subscription costs.
The other is artists. Spotify's mission is, roughly, a million artists living off their art and billions of fans enjoying it. Its Loud & Clear report (March 11, 2026) says more than 13,800 artists generated at least $100,000 from Spotify in 2025. Real progress, but those are royalties generated, paid to rights holders and split under contracts the report doesn't show, and far short of a million. That gap is the pain her bet answers.
To practice reading demand from public sources, try HubSpot Company Knowledge: Business Model, Competitors and Market Position.
Is it worth money?
Spotify reports two segments, Premium (subscriptions) and Ad-Supported. Hana copies key lines from its release for the quarter ended September 30, 2026 (published October 22, 2026), beside a year earlier:
| Q3 2026 | Q3 2025 | |
|---|---|---|
| Revenue | €4,969m | €4,272m |
| Premium revenue | €4,487m | €3,835m |
| Ad-Supported revenue | €482m | €437m |
| Gross margin | 33.0% | 31.6% |
| Monthly active users | 788m | 713m |
| Premium subscribers | 306m | 281m |
Subscriptions bring in about 90% of revenue (4,487 ÷ 4,969 ≈ 0.90). Ads are the lagging line, up 10% in a year against 17% for Premium. Gross margin sits near a third, though up 1.4 points, because most of what a stream costs goes to rights holders as royalties. That's the model's deepest weakness, and labels won't give up their share. Profit is recent too: 2024 was Spotify's first full year of net profit. Her position: live with it in music, and grow podcasts and audiobooks, where she suspects Spotify keeps more of each euro. Her bet would add money of the same kind, she hopes: a share of what fans pay an artist directly, not a royalty on a stream.
The skeptic's money objection: most of the 788 million monthly users are probably casual, so revenue may be small next to about €5 billion a quarter. She agrees. The point is keeping heavy listeners and artists, and she'd drop it if under 1% of a pilot artist's top listeners join within six months.
Two traps sit in these numbers. The 499 million ad-supported users plus 306 million subscribers make 805 million, not 788 million: the annual report counts them differently (monthly users listened in the last 30 days; subscribers include every paying account), so she never adds them. And Spotify restated past quarters in January 2026, so a figure from an older release won't match.
Read the earnings call for what management stressed and analysts pressed on. An AI assistant can gather it all, but check each number against the filing; assistants mix up quarters and invent figures.
Grow the ads, price the plans
Interviewers also ask two money questions that have nothing to do with your bet. The first is how you'd grow the money: pick the weakest line, say how you'd lift it, and name the figure that would move. Hana picks ads. About 499 million people hear them, yet they bring in under a tenth of revenue. She'd sell more ads inside podcasts, where listeners pay closer attention, and would call it working if Ad-Supported revenue outgrew Premium.
The second is pricing, and a question about the interviewer's own plans assumes you know the current ones. In the US since February 2026, Premium costs $12.99 for one person, $6.99 for a student, $18.99 for Duo (two accounts) and $21.99 for Family (up to six).
The step that worries Hana is graduation. A student moving to the individual plan goes from $6.99 to $12.99, 86% more, just as their life changes. She'd test a lower first-year price for graduates against a group without it, comparing cancellations over three months, and leave the headline price alone so soon after February's rise.
For a view on a company's prices, try Build a Pricing Strategy for Dropbox's Cloud Storage and File Sharing Plans. The method is in pricing and ad questions.
Who could block or copy it?
The main blocker is labels. They own most big artists' recordings and will want a share and a say. Hana would start with independent artists who own theirs, and widen only if the pilot works.
For copiers, she draws three rings of rivals, since the one that hurts most often sits outside the category.
| Ring | Examples | What they compete for |
|---|---|---|
| Same category | Apple Music, YouTube Music, Amazon Music | The subscription, often sold inside a bigger bundle |
| Substitutes | YouTube and TikTok for finding songs, car radio, audiobook stores | Listening minutes, and above all discovery |
| AI-native | Song-generating apps, and AI-made tracks on every service | Background listening, where few check who made it |
Against same-category rivals, Spotify stands apart on recommendations and on being nobody's side product. Its defense is that every listener's skips and saves sharpen everyone else's playlists, so it improves as Spotify grows. A rival could copy the Monday playlist, or a members page, in a week, but not years of listening by hundreds of millions of people, which is what tells an artist who their top listeners are. Against substitutes, Spotify has music, podcasts and audiobooks in one place. Against AI:
"An AI tool can make a decent focus track. It can't make someone buy a ticket for a singer they found on a Monday playlist. Spotify's defense is the bond between fans and artists. Its exposure is background listening, where nobody looks at the name."
That ring also holds her biggest worry: generated tracks and fake streams could crowd the catalog and drain the royalty pool, so she'd label AI tracks and catch suspicious streams before payouts.
For a rival behind a sales team, try Competitive Research on Enterprise Rivals Whose Products Sit Behind a Sales Process. For positioning against AI-native rivals, try Gemini Competitive Strategy: Finding a Defensible Edge Over Rival AI Assistants.
Competitive Research on Enterprise Rivals Whose Products Sit Behind a Sales Process
A product strategy interview question in which a PM at a B2B enterprise software company must analyze rivals whose products can only be seen through a...
Gemini Competitive Strategy: Finding a Defensible Edge Over Rival AI Assistants
Product strategy question on how Google should set Gemini apart from rival AI assistants such as ChatGPT, Claude and Microsoft Copilot.
Does it extend what Spotify is already doing?
Hana's claim: Spotify wants to be the one app for everything people listen to, because more formats bring more hours, and more hours mean fewer cancellations, which makes price rises survivable. Its own moves fit. Its third-quarter 2026 release says audiobooks will reach more than 180 markets, up from 22, by the end of 2026. US prices rose $1 to $2 a plan in February, and subscribers still grew 9% year on year in the third quarter, though US cancellations aren't broken out. Listening data shapes every screen, and the free tier keeps feeding the paid plans.
Her bet sits on the same line. Podcast memberships have launched in 34 markets, and creators earning from fans as well as ads are less likely to leave. At its May 2026 investor day, Spotify announced Reserved, which holds concert tickets for an artist's most devoted Premium listeners before general sale, starting in the US with Live Nation. Superfans now get early tickets inside Premium, so her membership must sell what Reserved can't: a regular connection to the artist.
One move cuts against the claim. The headline Premium plans now bundle audiobook hours while prices rise. A music-only Basic plan survives in the US only for some existing subscribers, so a new music-only listener pays for hours they never use. She'd raise the bundle as a question, not an accusation. Her claim runs from hours to fewer cancellations, so her graduate price aims at keeping listeners, and she judges her bet on loyalty: it gives a music-first fan a reason to stay that isn't more formats.
When the company is private or young
A private company publishes no earnings release, so build the picture from pieces: founders' interviews and podcasts for what it claims to build, job posts for what it's staffing, changelogs for what shipped and how often, and LinkedIn headcount or Glassdoor and Blind reviews for signs of growth or strain. Label usage figures from app-intelligence and web-traffic tools as rough estimates.
With no revenue yet, the money question for your bet becomes which model the company should test first, such as a paid team plan or a charge per use. Defend the order.
Pitch the bet with both objections
The night before, Hana checks that every line of her pitch points to a note in the brief, and that she raises the labels and the casual majority before a skeptic does. If she can't, the brief is still a pile of facts.
Show you share the company's priorities through how you'd judge the bet, not by echoing its mission back. Compare:
"I'm passionate about unlocking human creativity."
"I'd judge the pilot by how many mid-sized artists earn more from members than they made from streams over the same months."
The second makes artists earning a living the measure of success. The same brief feeds her why this company answer, her questions in culture and questions to ask, and her reading of the PM questions reported from Spotify.
To practice, answer Pitch Your Company's Next Big Bet When the CEO Stops You Over Lunch for your target company, with your bet and two objections in under a minute. Then try Uber CEO Exercise: Pick the Company's Next Strategic Priority and Defend It, starting from a short brief on Uber.
Uber CEO Exercise: Pick the Company's Next Strategic Priority and Defend It
A product strategy case that puts the candidate in the Uber CEO's seat and asks what the company should do next.
Pitch Your Company's Next Big Bet When the CEO Stops You Over Lunch
An open-ended product strategy question in which the CEO stops a product manager at lunch and asks what the company should do next.