What to expect, what they’re really testing, and what a strong answer looks like — scored.
Content discovery, personalization, engagement, and retention. Netflix PMs must understand the tension between content investment (expensive) and subscriber retention (the business metric), and think carefully about how recommendation systems shape user behavior.
The question below was asked by Netflix interviewers. The answer is graded on the five dimensions real PM interviewers use: structure, specificity, reasoning, decision quality, and delivery.
“Netflix is seeing a rise in subscription cancellations after the first month. What do you investigate?”
A 30-day cancellation spike tells me the problem is either an expectation gap (users signed up for something the product doesn't deliver) or a content exhaustion problem (users binged what they came for and see nothing else worth staying for). I'd investigate both.
First, I'd segment the cancelling users by what they did in their first 30 days: How many shows did they start? How many did they finish? Did they complete the show they signed up for? My hypothesis: a significant portion of month-1 cancellers signed up to watch one specific popular series, finished it, and found nothing else compelling enough to justify the subscription.
This is the 'one-show subscriber' problem — and it's addressable without spending on new content.
I'd pull data on: time to first 'content exhaustion' signal (user browses for 5+ minutes without starting anything), and what users in the control group (month-1 retainers) did differently in week 3-4. I'd expect retainers started a second, unrelated show before finishing their first one — they discovered a new content habit.
Based on this, I'd test two interventions: first, a 'while you watch' recommendation surface embedded in the player for episode 3-4 of popular limited series — before users finish the show they came for. Second, a 'week 3 check-in' notification that surfaces three shows in a genre the user hasn't tried, based on their watch history. Goal: get users to start a second habit before they lose one.
Success metric: 45-day retention rate (primary). I chose 45 over 30 because I want to see if we've created a second habit, not just deferred the cancellation.
Diagnoses before proposing, identifies the specific segment (one-show subscribers), and designs interventions around the behavior.
Names specific signals (5+ min browse without starting, episode 3-4 recommendation trigger) and a concrete hypothesis.
The 'one-show subscriber' insight is grounded in real Netflix behavior patterns and drives the intervention design directly.
Commits to two interventions with a clear causal logic; 45-day vs. 30-day metric choice is well-reasoned.
Good length; the 45-day metric justification adds real signal without padding.
The 'one-show subscriber' framing is the insight that makes this answer stand out — it's a real, well-documented Netflix retention problem and the answer builds directly from that hypothesis to the intervention. The episode 3-4 recommendation trigger is smart product thinking because it addresses the problem before it becomes a cancellation decision. The weakness is that the answer doesn't acknowledge the cost of more aggressive in-product recommendations — there's a risk of annoying users who are currently enjoying their show.
Add a guardrail metric (e.g., episode completion rate on the series they came for) to ensure the mid-show recommendations don't disrupt the viewing experience they signed up for.
Paste any Netflix PM interview question and your answer. Get scored on the same five dimensions — instantly, free, no signup.
Grade my answer free →First grade is free. No account needed.