Build a metric tree for Jobs after a major redesign

Problem Statement Description

Product context: LinkedIn is Microsoft's professional network; its products include profiles, feed, jobs, recruiting, LinkedIn Learning, sales tools, messaging, and ads.

LinkedIn has completed a major redesign of its Jobs experience, affecting how members discover roles, evaluate job details, decide whether to apply, and interact with post-application flows. The redesign may also influence employer outcomes, recruiter confidence, and how internal sales teams communicate value to hiring customers.

Your task is to define a metric tree that helps LinkedIn evaluate whether the redesigned Jobs experience is successful. The metric tree should connect the product’s top-level objective to measurable user behaviors across the job-seeker and employer marketplace, while making clear which metrics indicate progress, which diagnose friction, and which protect against unintended harm.

Focus on metric definition quality: what each metric means, who or what is counted, the denominator, where it is instrumented in the workflow, and how it would be segmented. The redesign should be evaluated not only by aggregate movement, but also by whether it improves outcomes for the right cohorts without degrading trust, relevance, or marketplace quality.

The experience should consider:

- The end-to-end Jobs funnel, from job discovery and search to job detail views, application intent, application completion, and post-application engagement.

- Marketplace balance between job seekers, employers, recruiters, and LinkedIn’s hiring/sales teams.

- Clear metric definitions, including numerator, denominator, time window, and event source.

- Cohorts such as new versus returning job seekers, active versus passive candidates, geography, seniority, industry, job type, and employer segment.

- Instrumentation needs for redesigned surfaces, including search, recommendations, job cards, job details, alerts, saved jobs, and apply flows.

- Guardrail metrics for relevance, spam, duplicate jobs, low-quality applications, member trust, employer satisfaction, and notification fatigue.

- Decision usefulness: which metrics determine launch success, which explain why performance moved, and which trigger further investigation.

The goal is to present a structured metric tree that a LinkedIn product team could use after launch to monitor redesign impact, diagnose user or marketplace friction, and make confident product decisions without relying on vanity metrics alone.