Diagnose Netflix's Shrinking Renewals: Root-Cause Hypotheses and How to Test Them

You are a product manager at Netflix. The share of subscribers who renew has been dropping compared with the same period in earlier years. What hypotheses could explain the drop, and how would you test each one to find the real cause? ```hint Pin down the metric first Before listing causes, decide exactly what "renewal" means for a subscription that mostly bills monthly, who sits in the denominator, and whether the drop could come from a change in who is subscribing rather than from any group becoming less loyal. ``` ### Clarifying Questions - How is the renewal rate defined: the share of members due to be billed in a period who pay and stay, or the share of a signup cohort still subscribed a year later? - Is the decline global, or concentrated in particular countries, plans (including the ad-supported plan), devices or signup cohorts? - When did the decline start, and is it gradual or a step change around an event such as a price change or a change to account-sharing rules? - Do downgrades to a cheaper plan count as renewals, and are cancellations caused by failed payments included? ### What a Strong Answer Covers - A precise definition of the renewal metric and a check that the decline is real rather than a change in measurement, billing or the mix of members - Segmentation by market, plan, tenure, acquisition source, device and voluntary versus involuntary churn to localize the decline before guessing at causes - A structured hypothesis tree (member mix, value delivered, price and plans, payment and product friction, competition and external factors) tied to Netflix-specific mechanisms - For each hypothesis, the data, experiment or research that would confirm or rule it out, checked in a sensible order from cheap to costly - How a confirmed cause turns into an action and a metric to watch ### Follow-up Questions - Renewal is down but the total number of paying members is up. How do you reconcile the two, and should leadership worry? - Your analysis shows cancellations cluster right after a popular series ends. What would you change, and how would you measure whether it worked? - How would you separate the effect of a price increase from the effect of a weaker release slate in the same quarter?