How would you detect unhealthy growth in Prime

Problem Statement Description

Product context: Amazon is a commerce, logistics, media, devices, and cloud company; its products include Marketplace, Prime, Prime Video, Alexa devices, ads, fulfillment, and AWS.

Amazon Prime is a multi-benefit membership spanning shopping, delivery, video, music, devices, and partner offers. Growth in Prime can look positive at the top level—more sign-ups, more members, higher engagement—but still be unhealthy if it is driven by low-quality acquisition, unsustainable costs, poor member experience, subsidy-heavy behavior, or short-term promotions that do not translate into durable customer value.

In this metrics interview, define how you would detect whether Prime growth is healthy or unhealthy. Focus on the member lifecycle from acquisition to activation, benefit usage, renewal, retention, and long-term value. Consider that Prime growth affects multiple Amazon systems: commerce demand, logistics capacity, media consumption, customer support, subscription revenue, and overall marketplace economics.

Your task is not to design a new Prime feature, but to propose a metric framework that helps Amazon distinguish durable, customer-obsessed growth from growth that may damage profitability, operations, trust, or long-term retention.

The experience should consider:

- What “growth” means for Prime: new members, active members, paid members, renewals, engagement, or revenue contribution.

- How to define unhealthy growth using clear numerators, denominators, and time windows.

- Which cohorts matter, such as new vs. existing members, trial vs. paid, monthly vs. annual, geography, acquisition channel, promotion type, and benefit usage pattern.

- What instrumentation is needed across sign-up, payment, cancellation, benefit usage, delivery experience, customer support, and renewal events.

- Which leading and lagging indicators would help detect problems before they appear in churn or financial results.

- What guardrail metrics should protect customer experience, logistics quality, contribution margin, abuse/fraud, and long-term trust.

- How to separate normal seasonal behavior, campaign effects, and mix shifts from genuine unhealthy growth.

- How the metrics would be useful for decisions such as scaling campaigns, changing eligibility, adjusting benefits, or investigating specific cohorts.

The goal is to build a practical measurement approach that helps Amazon evaluate Prime growth quality, diagnose where unhealthy patterns are emerging, and support better decisions without relying only on headline membership growth.