Company preparation
Company tracks. Know the loop.
Choose your route.
Every company weights product sense, execution, metrics, and leadership differently. Train for the sequence you will actually face.
Interview routes
Start with the company in sight.
Each track preserves the real question count, average difficulty, and round mix.
Google's PM loop leans hard on product sense with a strong analytical bar — expect 'improve X' plus a metrics deep-dive and one estimation.
Meta interviews are execution-heavy: expect metric drops to diagnose, goal-setting for social products, and a sense round on engagement loops.
Amazon anchors on Leadership Principles — every answer needs a STAR story with numbers, plus working-backwards product thinking.
Stripe tests API-product thinking and craft: developer empathy, precision in writing, and strategy for platform businesses.
Microsoft blends classic design questions with customer-focus and collaboration stories — breadth over trickiness.
Uber leans on marketplace dynamics: supply/demand balance, ops-heavy execution cases, and estimation grounded in real-world logistics.
Apple probes taste and restraint: what you would refuse to ship, how hardware and software constrain each other, and privacy as a product decision rather than a policy footnote.
YouTube is a three-sided problem: viewers, creators and advertisers pull in different directions. Expect recommendation tradeoffs and watch-time metrics that can be gamed.
Instagram interviews centre on engagement loops and creator supply: why a format spreads, what kills reach quality, and how ranking changes ripple through behaviour.
WhatsApp is deliberately minimal at enormous scale. Expect questions about what NOT to add, encryption constraints on features, and monetizing without breaking trust.
TikTok tests fast-loop thinking: cold-start recommendation, creator retention, and how a feed decision changes content supply within days rather than quarters.
Notion asks about flexible primitives: how a general building block serves opposing use cases, where templates beat features, and when bottom-up adoption turns into an enterprise sale.
Canva centres on non-expert users: removing intimidation from design work, template economics, and the jump from a single creator to a paying team.
OpenAI probes judgment under genuine uncertainty: shipping when capability is unreliable, designing for graceful failure, and weighing safety against usefulness with no clean answer.
Salesforce is enterprise-first: multi-stakeholder buying, admin vs end-user needs, and platform extensibility where your customer's customer also builds on you.
Adobe balances professional depth against new entrants: serving experts who resent change, pricing a creative suite, and folding generative tools into established workflows.
Atlassian tests team-workflow reasoning and its self-serve motion: configurability that does not become chaos, and expansion inside an org without a traditional sales push.
Dropbox questions focus on sync reliability as the product, differentiating a commoditized primitive, and converting individual storage habits into paid team collaboration.
Pinterest sits between inspiration and intent: taste-based discovery, shopping conversion without wrecking the browsing feel, and content supply for narrow interests.
Snap centres on close-friend communication and camera-first design: ephemerality as a feature, AR utility beyond novelty, and holding a young audience against larger rivals.
DoorDash is a three-sided marketplace with a physical clock: consumers, merchants and Dashers all break differently, and delivery-time promises make failure immediately visible.
Instacart hinges on catalogue accuracy and substitutions: an out-of-stock item is a product problem, not just an ops one, and retailer relationships shape what you can build.
Lyft asks marketplace-balance questions with a sharper focus on driver experience and differentiation as the second mover in a duopoly.
PayPal tests payments fluency: trust and fraud as product surfaces, two-sided checkout adoption, and designing inside regulatory limits rather than around them.
Discord centres on community mechanics: what makes a server stick, moderation as core product, and monetizing without taxing the communities that create the value.
Reddit is governed by its communities: ranking changes shift norms, moderators are unpaid stakeholders with real power, and growth can actively damage what makes a subreddit good.
Zoom rewards reliability-first thinking: quality under bad networks, frictionless joining, and expanding past the meeting without bloating the one thing that works.
Coinbase pairs consumer simplicity with irreversible actions: onboarding novices safely, designing when a mistake cannot be undone, and shipping under regulatory uncertainty.
PhonePe is built on UPI-scale payments in India: designing for first-time smartphone users, low-connectivity reliability, and monetizing on top of near-zero-margin rails.