Build a metric tree for Reservations after a major redesign

Problem Statement Description

Product context: Uber is a mobility and delivery platform; its products include rides, Uber Eats, grocery and retail delivery, freight, driver and courier tools, and marketplace pricing.

Uber has recently completed a major redesign of its Reservations experience for couriers. The redesign may affect how couriers discover reserved opportunities, evaluate them, commit to them, prepare for them, and successfully complete the reserved work within a marketplace that depends on reliability, liquidity, and operational trust.

Your task is to build a metric tree that helps Uber understand whether the redesigned Reservations experience is working. The metric tree should connect the product’s business and marketplace goals to measurable courier behaviors, user experience signals, and operational outcomes. It should be useful for diagnosing where performance is improving or degrading after launch.

Assume this is a global product surface with variation across cities, courier types, delivery categories, supply conditions, and local operating rules. The metric framework should be clear enough for product, data science, operations, and engineering teams to use in post-launch monitoring and decision-making.

Your metric tree should consider:

- A clear top-level success metric and the denominator it is measured against

- Funnel stages across discovery, reservation selection, commitment, preparation, arrival, and completion

- Courier cohorts such as new vs. experienced couriers, high-frequency vs. occasional couriers, and market-level differences

- Instrumentation needed to distinguish exposure, intent, action, cancellation, no-show, and completion events

- Marketplace guardrails around customer reliability, courier earnings, supply availability, cancellations, and operational costs

- Quality and trust signals that capture whether reservations are predictable, understandable, and fair for couriers

- How the metric tree would support diagnosis if the redesign improves one part of the funnel but harms another

The goal is to create a structured measurement framework that Uber could use to evaluate the redesigned Reservations product, monitor launch health, and identify the most important areas for follow-up investigation.