Design an A/B test to improve verified first financial action rate for new investors and salary earners

Problem Statement Description

You are evaluating an onboarding experiment for a consumer-finance product serving new investors and salary earners. The business wants more new users to complete a “verified first financial action,” such as making their first compliant investment, setting up a savings action, completing a salary-linked money movement, or taking another regulated financial step after KYC and eligibility checks.

The current onboarding flow may create friction around trust, documentation, risk disclosures, payment setup, suitability, or clarity of what action to take first. Your task is to design an A/B test that can determine whether a proposed onboarding or activation change improves this verified first action rate without creating compliance, fraud, customer harm, or poor-quality activation issues.

Focus on how you would define the success metric, who should be included in the experiment, how the event should be instrumented, and how the result would be interpreted. The test should be suitable for a regulated consumer-finance environment where privacy, consent, financial suitability, and operational reliability matter.

The experience should consider:

- A precise definition of “verified first financial action,” including what qualifies and what should be excluded.

- The denominator for the primary metric, such as newly onboarded users, KYC-completed users, eligible users, or users exposed to the experiment.

- Key user cohorts, including new investors versus salary earners, KYC status, risk profile, income band, acquisition channel, and platform.

- Instrumentation needed to track exposure, eligibility, funnel steps, verification status, failed attempts, and downstream completion.

- Guardrail metrics such as complaint rate, failed KYC/payment rate, fraud signals, cancellation or reversal rate, support contacts, suitability violations, and retention.

- Experiment design choices including randomization unit, sample eligibility, duration, power, novelty effects, and handling users who cross segments.

- Decision usefulness: how you would determine whether the experiment result is strong enough to launch, iterate, or stop.

Your goal is to frame a clear A/B test plan that helps the product team make a trustworthy decision about improving activation for new finance users while maintaining regulatory compliance, customer trust, and quality of financial outcomes.