Determine Metrics to Measure Free-Trial Impact on Subscriptions

Quick Overview

Determine Metrics to Measure Free-Trial Impact on Subscriptions evaluates metric design, causal reasoning, experiment setup, diagnostics, SQL/statistical checks, and recommendations in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

Determine Metrics to Measure Free-Trial Impact on Subscriptions

Company: OpenAI

Role: Data Scientist

Category: Analytics & Experimentation

Difficulty: medium

Interview Round: Take-home Project

##### Scenario A/B test that offers free users a free trial of the paid plan; dataset includes user_id, treatment assignment, sign-ups and cancels. ##### Question Which primary metrics would you track to measure the free-trial’s impact on subscription behaviour? Define each metric precisely. Calculate the lift in conversion to paid subscription between treatment and control and test whether the difference is statistically significant. Estimate the treatment effect on churn (cancels) within 30 days. Is the effect practically meaningful? Given your results, would you recommend rolling the free-trial out to all users? Justify your answer. ##### Hints Choose consistent denominators, run two-sample tests, and discuss both statistical and business significance.

Quick Answer: Determine Metrics to Measure Free-Trial Impact on Subscriptions evaluates metric design, causal reasoning, experiment setup, diagnostics, SQL/statistical checks, and recommendations in a realistic interview setting. A strong answer states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

|Home/Analytics & Experimentation/OpenAI
OpenAI logo
OpenAI
Aug 4, 2025, 10:55 AM
mediumData ScientistTake-home ProjectAnalytics & Experimentation
60
0

Determine Metrics to Measure Free-Trial Impact on Subscriptions

A/B Test: Free Trial Offer Impact on Subscription Behavior

Scenario

You are analyzing a randomized A/B test in which free users are offered a free trial of the paid plan (treatment) versus no trial offer (control). Each user is assigned exactly once at first exposure.

Available Data (assumed minimal schema)

  • user_id
  • variant: "treatment" or "control"
  • assigned_at: timestamp of assignment/exposure
  • subscribed_at: timestamp when a user becomes a paid subscriber (null if never subscribes)
  • canceled_at: timestamp when a user cancels their paid subscription (null if never cancels)

Assume you have at least 30 days of observation after assignment for all users included in the analysis (or you censor/exclude users without a full window consistently across arms).

Tasks

  1. Define the primary metrics you would track to measure the free-trial’s impact on subscription behavior. Use precise, unambiguous definitions with clear numerators, denominators, and time windows.
  2. Calculate the lift in conversion to paid subscription between treatment and control.
  3. Test whether the difference in conversion is statistically significant (state the test and how you’d compute it).
  4. Estimate the treatment effect on churn (cancels) within 30 days and discuss whether the effect is practically meaningful.
  5. Based on your results, would you recommend rolling the free-trial out to all users? Justify your answer considering both statistical and business significance.

Hints

  • Use consistent denominators (intent-to-treat where appropriate).
  • Use two-sample tests with appropriate confidence intervals.
  • Discuss both statistical and practical/business significance.

Clarifying Questions to Ask Guidance

  • Clarify the business objective, unit of analysis, time window, exposure definition, and primary metric.
  • State assumptions about instrumentation, randomization, sample size, and data quality.
  • Separate descriptive analysis from causal claims.

What a Strong Answer Covers Guidance

  • A metric framework with primary, guardrail, and diagnostic metrics.
  • A credible analysis or experiment design with clear assumptions and bias checks.
  • SQL/statistical logic for segmentation, variance, confidence, and data validation where relevant.
  • An actionable recommendation that explains trade-offs and next steps.

Follow-up Questions Guidance

  • What sanity checks would you run before trusting the result?
  • How would you handle novelty effects, seasonality, or selection bias?
  • What decision would you make if metrics disagree?
Loading comments...