Dataford
Popular roles
Software EngineerData AnalystData ScientistData EngineerBusiness AnalystAI EngineerMachine Learning EngineerProduct Manager
Browse
Browse All RolesEvery role hub, from analyst to MLBrowse All CompaniesCompany-specific interview loopsAll Interview GuidesThe full guide library
Top questions by role
Software EngineerData AnalystData ScientistData EngineerBusiness AnalystAI EngineerMachine Learning EngineerProduct Manager
Top questions by skill
SQLPythonStatisticsMachine LearningA/B TestingSystem DesignGenerative AIProduct SenseMetricsBehavioral
Browse all questions →Try a mock interview
ExperiencesCoaching
New
Practice
Mock InterviewsTimed interview simulations with feedbackSuccess PathYour 6-week structured planModulesCurated lessons by topicWebinarsTalks from ex-Big Tech data leadsPlaygroundA free-form scratch editor
Learn
BlogInterview strategy and career adviceTech Job Market ReportHiring trends across data and AI rolesFor UniversitiesDataford for career centersAbout DatafordWho we are and how we build
Pricing
Get started

A/B Test for Checkout Conversion Quality

HardStatistics & Probability00:00
Practice interviewer
In session
5 left
00:00

Whenever you're ready, walk me through how you'd approach this.

You need to log in / sign up to chat or submit.

Problem

Business Context

Affirm is testing a redesigned checkout entry experience on the Affirm checkout surface to increase application-start-to-checkout conversion. Product wants to ship only if the change improves top-of-funnel conversion without degrading downstream quality, measured by 30-day repayment delinquency among originated loans.

Problem Statement

Design and analyze the experiment using the data below. Determine whether the treatment should be rolled out if it increases checkout conversion but must not materially worsen downstream credit quality.

Given Data

The experiment ran for 21 days with user-level randomization on eligible checkout sessions.

MetricControlTreatment
Eligible checkout sessions120000120000
Completed checkouts1440015120
Conversion rate12.00%12.60%
Originated loans1008010584
30-day delinquent loans252296
Delinquency rate among originated loans2.50%2.80%

Additional test settings:

ParameterValue
Significance level for conversion test0.05
One-sided non-inferiority margin for delinquency0.003
Minimum detectable lift planned for conversion0.005
Desired power for conversion test0.80

Requirements

  1. Define the primary metric, guardrail metric, and the null/alternative hypotheses for both.
  2. Test whether conversion improved using a two-proportion z-test.
  3. Construct a 95% confidence interval for the conversion lift.
  4. Test whether delinquency is non-inferior using the stated margin.
  5. State the launch decision under the rule: ship only if conversion is significantly higher and delinquency is not worse than the margin.
  6. Briefly comment on whether the sample size appears adequate for the planned conversion MDE.

Assumptions

  • Randomization is valid and each eligible session appears once.
  • Delinquency is evaluated only on originated loans.
  • No major traffic-mix shift occurred during the 21-day window.
  • Ignore multiple secondary metrics beyond the stated guardrail.