I’ve been switching between top-down and bottom-up approaches depending on prompt clarity. One vet told me: use top-down when the interviewer gives broad market signals or industry figures; use bottom-up when user behavior or unit economics are front-and-center. In a recent mock I started top-down, then pivoted to a quick bottom-up sanity check and that combination saved me when interviewers probed further. How do you decide which route first in ambiguous prompts?
here’s the blunt answer: pick whichever gives you a defensible headline faster. top-down looks polished if the market is well-defined; bottom-up is safer for product-oriented questions. don’t try to impress with both unless you have spare time. the real trick is to show you can triangulate quickly—say ‘i’ll start top-down for speed, then check with a bottom-up if time permits.’ that line sounds deliberate and smart.
in interviews people often flip to bottom-up because it feels more ‘practical.’ pity they then invent unrealistic unit economics. vets prefer the approach that reveals your weakest assumption early. choose the one that surfaces a testable assumption in the first minute.
i usually go top-down first because it gives me a quick number to reference, then i do a tiny bottom-up check if i have time. helps me stay organized and not lost.
tried bottom-up today for a product case and it felt more natural. still learning when to switch tho.
The deciding factors are the question’s framing and the data points the interviewer offers. If they drop industry revenue or market share cues, top-down is efficient. If user metrics, pricing tiers, or unit economics are given, build bottom-up to surface operational levers. Teach candidates to articulate the reason for their chosen approach: that meta-commentary demonstrates strategic judgment and allows the interviewer to guide you if they expected a different angle.
great instincts! mix both when stuck — it’s a smart way to show flexibility and rigour. keep it up!
another time a vet recommended ‘start top-down but don’t bank it—use bottom-up as your confidence check.’ acting on that saved me when my initial market estimate seemed off; the quick bottom-up exposed the overestimate and let me correct mid-answer.
In practice, binary rules help: use top-down when at least one reliable macro input exists (industry revenue, population), and use bottom-up when you can credibly model user behavior with 2–3 parameters (users, frequency, spend). I tracked five mocks and found that starting top-down reduced total case time by 15% while still allowing a bottom-up plausibility check when needed—an efficient hybrid strategy.
If you must choose, pick the approach that produces a falsifiable output quickest. For top-down, that’s market revenue; for bottom-up, it’s revenue per user. Explicitly call out which variable you’d stress-test first. Interviewers pay attention to how you prioritize sensitivity, not just the model you build.