How would you detect unhealthy growth in Treasury

Problem Statement Description

Product context: Stripe is financial infrastructure for internet businesses; its products include payments, Checkout, Billing, Connect, Radar, Issuing, Terminal, and tax tools.

Stripe Treasury enables startups and platforms to offer or use financial-account-like capabilities as part of their business workflows, including storing funds, moving money, and supporting operational finance needs. Growth in this product can look strong on the surface, but may be unhealthy if it comes from low-quality usage, risky money movement, poor retention, operational strain, regulatory exposure, or customers whose activity does not translate into durable value.

In this metrics interview, you are being asked to define how Stripe should detect whether Treasury growth is healthy or unhealthy. The focus is not just top-line adoption, but whether the growth is sustainable, compliant, reliable, and valuable for both Stripe and its startup customers.

Assume you are evaluating a live Treasury product with multiple customer cohorts, varying transaction patterns, and financial-risk constraints. Your task is to describe what you would measure, how you would instrument it, how you would segment the data, and how those metrics would help product, risk, operations, and leadership make decisions.

The experience should consider:

- What “growth” means for Treasury, including the unit of measurement and the denominator used for each metric

- How to distinguish account creation, activation, funded usage, repeat usage, and meaningful retained usage

- Which customer cohorts, segments, geographies, use cases, or acquisition channels may show different quality patterns

- How to detect growth driven by fraud, compliance risk, excessive support burden, operational failures, or concentration risk

- What instrumentation is needed across onboarding, KYC/KYB, account funding, money movement, balances, failed transactions, disputes, and support events

- Which guardrail metrics would prevent over-optimizing for volume, balances, or customer count alone

- How to separate short-term spikes, seasonality, customer mix changes, and product changes from true unhealthy growth

- How the metric system would support decisions such as investigation, throttling, product changes, risk review, or go-to-market adjustments

The goal is to design a metrics approach that gives Stripe an early-warning system for Treasury growth quality, helping teams understand not only whether the product is growing, but whether that growth is durable, safe, compliant, and aligned with long-term customer and business value.