← Chime Interview Insights

Chime·Data Scientist·Technical Phone Screen·Senior

SeniorPrefer not to say
May 2026Remote

Summary

Chime data scientist interview with a single massive case study built around a messy A/B test for a fitness tracker launch. The whole thing is one long scenario with six sub-questions layered on top of each other, which felt more like a take-home than a live interview. Dense but fair if you've done experimentation work before.

Questions Asked (6)

Q1

For a user-level A/B test with overlapping marketing activity across channels, what randomization unit and exposure rule would you use to limit contamination? How do you handle users who receive ads but were never formally randomized, or who end up in both variants across different platforms?

A/B Testing & ExperimentationTechnical Trade-offs
Author's notes

This is where I spent the most time and probably said too much.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the randomization unit and exposure rule that minimize contamination given overlapping marketing channels, such as randomizing at the user level and defining exposure as ad view. Then address edge cases like users exposed without randomization and cross-platform contamination by using intent-to-treat or instrumental variables, and discuss trade-offs.

Pro tip: Acknowledge that perfect isolation is impossible in multi-channel marketing; instead, focus on quantifying and bounding contamination's impact, and propose sensitivity analyses to validate results.

1. Choose Randomization Unit

Select a unit that aligns with the intervention and minimizes spillover, such as user-level for cross-device campaigns, and justify why it's appropriate for the marketing channels involved.

2. Define Exposure Rule

Specify what constitutes exposure (e.g., ad view, click) and how it's measured consistently across channels, ensuring it reflects the intended treatment.

3. Handle Unrandomized Exposed Users

Decide whether to include them in the analysis (e.g., via intent-to-treat) or exclude them, and discuss the bias-variance trade-off and potential use of instrumental variables.

4. Address Cross-Platform Contamination

Detect and mitigate users appearing in both variants across platforms by using unique user IDs, cross-device graphs, or probabilistic matching, and consider excluding or adjusting for them.

5. Validate and Quantify Contamination

Run diagnostics like checking for pre-treatment differences, using placebo tests, or simulating contamination to estimate its impact on effect estimates.

Key Points to Mention

  • Randomization unit options: user, device, cookie, and their trade-offs in multi-channel settings.
  • Exposure definition: intent-to-treat vs. treatment-on-treated, and the importance of consistent measurement.
  • Contamination sources: spillover, cross-device usage, and overlapping campaigns.
  • Handling unrandomized exposed users: include via ITT, exclude, or use instrumental variables.
  • Cross-platform contamination: use of universal user IDs, device graphs, and deduplication.
  • Sensitivity analyses: quantifying contamination bias and robustness checks.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Define a primary success metric and at least three guardrail metrics for this experiment, including how each is computed, the measurement window, and any exclusion rules.

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

Went with 14-day purchase conversion as primary since it matched the stated business goal and the pre-period baseline gave a clean reference point.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the product surface (e.g., a new onboarding flow or a feature in the Chime app). Then define a primary success metric that directly measures the intended impact, and select guardrail metrics that protect against unintended harm to user experience, business health, and long-term retention. For each metric, specify the exact computation (numerator/denominator), measurement window (e.g., 7-day or 28-day), and exclusion rules (e.g., internal users, bots, or users with pre-existing conditions).

Pro tip: Tie guardrails to Chime's specific business model—e.g., deposit frequency, direct deposit setup, or fee avoidance—and mention that you'd pre-register the metrics and analysis plan to avoid p-hacking and ensure statistical rigor.

1. Clarify the experiment and hypothesis

Restate the experiment's goal and the expected user behavior change. Identify the primary decision the experiment will inform (e.g., ship, iterate, or kill).

2. Define the primary success metric

Choose a metric that directly measures the intended impact and is sensitive enough to detect a meaningful effect. Specify its computation, measurement window, and exclusion rules.

3. Select guardrail metrics

Pick at least three guardrails covering different risk areas: user experience (e.g., support contacts), business health (e.g., revenue or engagement), and long-term retention. For each, define computation, window, and exclusions.

4. Specify measurement windows and exclusions

For each metric, state the time frame (e.g., 7-day, 28-day) and any exclusion criteria (e.g., internal employees, bots, users with incomplete data). Explain why these choices are appropriate.

5. Summarize and link to decision-making

Recap how the primary and guardrail metrics together will guide the ship/no-ship decision, and mention any statistical considerations (e.g., power, multiple comparisons).

Key Points to Mention

  • Primary metric should be directly tied to the experiment's hypothesis and business objective (e.g., increase in activated users).
  • Guardrail metrics must cover different dimensions: user experience (e.g., complaint rate), business (e.g., revenue per user), and long-term health (e.g., 28-day retention).
  • Computation details: clearly define numerator and denominator, and whether it's a ratio, average, or count.
  • Measurement window: choose a window that captures the full effect without introducing noise (e.g., 7 days for short-term, 28 days for retention).
  • Exclusion rules: exclude internal users, bots, and users with pre-existing conditions that could skew results.
  • Pre-registration and multiple testing correction: mention that metrics and analysis plan are pre-registered, and consider corrections like Bonferroni or false discovery rate.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q3

What specific diagnostics would you run to check for sample ratio mismatch, instrumentation gaps, novelty effects, and day-of-week seasonality? For each, what test or threshold would you use, and what would you do if it fails?

A/B Testing & ExperimentationRoot Cause Analysis
Author's notes

SRM check via chi-square on assignment counts, threshold at p < 0.01.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer by addressing each diagnostic separately: sample ratio mismatch (SRM), instrumentation gaps, novelty effects, and day-of-week seasonality. For each, describe the specific test or threshold you would use and the remediation steps if it fails. Emphasize a systematic, data-driven approach that prioritizes actionable insights.

Pro tip: Always pre-register your diagnostics and thresholds before the experiment starts to avoid p-hacking and ensure objectivity. When an issue is detected, focus on root cause analysis rather than just fixing symptoms.

1. Sample Ratio Mismatch (SRM)

Run a chi-squared goodness-of-fit test to compare observed vs. expected traffic split. Use a threshold of p < 0.001 or a relative difference > 2%. If SRM is detected, investigate logging, randomization, or filtering issues and consider invalidating the experiment.

2. Instrumentation Gaps

Check for missing data, unexpected nulls, or discrepancies between event counts and expected volumes. Use data quality dashboards, anomaly detection, or SQL audits. If gaps exist, trace the data pipeline, fix instrumentation, and re-run the experiment if necessary.

3. Novelty Effects

Analyze treatment effect over time (e.g., daily or weekly) to see if it diminishes. Use a t-test or regression with time interaction, or compare early vs. later periods. If novelty is present, extend the experiment or focus on long-term metrics.

4. Day-of-Week Seasonality

Compare metrics across days of the week using ANOVA or visual inspection. Check if treatment effects vary by day. If seasonality is an issue, ensure balanced randomization across days or use stratified analysis.

5. Remediation and Decision

For each failure, document the issue, assess impact on validity, and decide whether to fix, re-run, or abandon the experiment. Communicate findings to stakeholders and update best practices.

Key Points to Mention

  • Chi-squared test for SRM with p < 0.001 threshold
  • Data quality checks: nulls, duplicates, event volume anomalies
  • Time-series analysis for novelty effects (e.g., daily treatment effect trends)
  • ANOVA or day-of-week stratification for seasonality
  • Root cause analysis and remediation steps for each diagnostic
  • Pre-registration of diagnostics to avoid p-hacking

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q4

How would you adjust for the concurrent marketing push that overlapped the test period? Walk through at least one concrete method, and explain the trade-offs versus alternatives.

A/B Testing & ExperimentationTechnical Trade-offsAdaptability & Ambiguity
Author's notes

I led with CUPED using pre-period engagement as the control variate, since we had 14-day baselines and it's computationally straightforward.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the nature of the concurrent marketing push and its potential impact on the experiment. Then, propose a concrete adjustment method such as using a difference-in-differences approach or including marketing exposure as a covariate, and discuss the trade-offs compared to alternatives like re-randomization or excluding affected users.

Pro tip: Acknowledge that perfect isolation is rare in real-world experiments; demonstrating how you'd quantify and mitigate bias while maintaining statistical power shows maturity. Also, mention the importance of pre-registering the adjustment plan to avoid p-hacking concerns.

1. Clarify the overlap and impact

Ask clarifying questions about the marketing push: timing, target audience, and whether it was randomized. Assess how it might differentially affect treatment and control groups.

2. Choose an adjustment method

Select a concrete method like difference-in-differences (if pre-period data exists), propensity score weighting, or including marketing exposure as a covariate in your model.

3. Implement and validate

Apply the method, check for balance after adjustment, and run sensitivity analyses to ensure results are robust to different assumptions.

4. Discuss trade-offs

Compare your chosen method to alternatives (e.g., excluding affected users, re-randomizing, or ignoring the issue) in terms of bias, variance, and practicality.

5. Conclude with recommendation

Summarize why your approach balances validity and business constraints, and suggest how to prevent similar issues in future experiments.

Key Points to Mention

  • Difference-in-differences (DiD) to control for time-varying confounders
  • Propensity score matching or weighting to balance marketing exposure
  • Including marketing exposure as a covariate in regression adjustment
  • Trade-offs: bias reduction vs. loss of sample size or statistical power
  • Assumption checks: parallel trends for DiD, overlap for propensity scores
  • Pre-registration of analysis plan to maintain integrity

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q5

With a 6% baseline conversion rate, a 50/50 split, alpha of 0.05 two-sided, 80% power, and roughly 2.8 million eligible users over the test window, what is the minimum detectable relative lift? Is the experiment adequately powered?

A/B Testing & ExperimentationProduct Analytics & Metrics
Author's notes

I worked through the binomial variance formula live and got to an absolute MDE somewhere around 0.3 to 0.4 percentage points, which translates to roughly a 5 to 7 percent relative lift at 80% power.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, compute the minimum detectable effect (MDE) for a two-proportion z-test using the given parameters: baseline conversion rate (p=0.06), alpha=0.05 (two-sided), power=0.80, and equal allocation. Then, compare the required sample size per variant to the available sample size (2.8M total, so 1.4M per variant) to determine if the experiment is adequately powered. Finally, interpret the MDE in relative terms and discuss practical implications.

Pro tip: Always clarify whether the MDE is absolute or relative, and state the assumptions (e.g., normality, independence). Also, mention that with such a large sample, even tiny lifts become detectable, so consider practical significance, not just statistical significance.

1. Identify parameters and formula

List the given values: baseline conversion rate p=0.06, alpha=0.05 (two-sided), power=0.80, allocation ratio=1:1, total sample size N=2.8M. Use the standard formula for sample size per variant for a two-proportion z-test: n = ( (z_{1-alpha/2} + z_{1-beta})^2 * (p1(1-p1) + p2(1-p2)) ) / (p2 - p1)^2, where p2 = p1 + delta.

2. Compute MDE given available sample

Since total sample is fixed, solve for the minimum detectable absolute difference delta (p2 - p1) that yields power 0.80 with n=1.4M per variant. Use the formula: delta = (z_{1-alpha/2} + z_{1-beta}) * sqrt( (p1(1-p1) + p2(1-p2)) / n ). Because p2 depends on delta, iterate or approximate by using p1(1-p1) for both groups (conservative).

3. Calculate relative lift

Convert the absolute MDE to relative lift: relative MDE = delta / p1. For example, if delta ≈ 0.0005, then relative lift ≈ 0.0005/0.06 ≈ 0.83%.

4. Assess power adequacy

Compare the computed MDE to the effect size you care about. If the MDE is smaller than the smallest meaningful effect, the experiment is adequately powered. With 2.8M users, the MDE will be very small (likely <1% relative), so the experiment is highly powered for typical lifts.

5. Discuss practical considerations

Mention that statistical significance does not imply practical significance. Also, note assumptions: independent observations, no interference, and that the baseline rate is accurate. Consider multiple testing if multiple metrics are evaluated.

Key Points to Mention

  • Use of two-proportion z-test for power analysis.
  • Formula for sample size and MDE, including z-scores for alpha and beta.
  • Conversion of absolute MDE to relative lift.
  • Comparison of required sample size to available sample size.
  • Interpretation: with 2.8M users, the experiment is overpowered for most practical effects.
  • Caveats: assumptions of independence, no peeking, and practical significance.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q6

After all your adjustments, the estimated lift in 14-day conversion is +3.5% with a 95% CI of -0.5% to +7.5%, ARPU is up 1.2%, and refund rate increased by 1.1 percentage points. Do you recommend launch, rollback, or an extended test? What thresholds drive that decision and how do you communicate it?

A/B Testing & ExperimentationStakeholder ManagementCross-functional Alignment
Author's notes

The CI crossing zero was the sticking point for me.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by acknowledging the mixed results and the uncertainty in the primary metric, then propose an extended test to gather more data, because the confidence interval includes zero and the refund rate increase is concerning. Explain the decision thresholds you would use, such as statistical significance, practical significance, and guardrail metrics, and how you would communicate the recommendation to stakeholders.

Pro tip: Emphasize that the refund rate increase is a red flag that could erode long-term value, so even if conversion lift becomes significant, you'd need to investigate the root cause before launching. Also, mention that you'd align with product and finance on the acceptable trade-off between conversion and refunds.

1. Assess statistical significance and uncertainty

Note that the 95% CI for conversion lift includes zero, so the result is not statistically significant. The ARPU increase is positive but small, and the refund rate increase is a potential guardrail violation.

2. Evaluate practical significance and business impact

Consider whether a +3.5% lift in conversion is meaningful for Chime, and weigh it against the 1.1pp refund rate increase. Calculate the net impact on revenue and customer lifetime value.

3. Check guardrail metrics and long-term effects

Refund rate is a key guardrail; an increase suggests potential customer dissatisfaction or product issues. Also consider other guardrails like churn, support tickets, or regulatory concerns.

4. Decide on action: launch, rollback, or extend test

Given the uncertainty and guardrail concern, recommend an extended test to gather more data, possibly with a larger sample size or longer duration, and investigate the refund rate increase.

5. Communicate recommendation to stakeholders

Present the trade-offs clearly, using non-technical language, and propose next steps: extend test, dive deeper into refunds, and set a decision deadline. Align with product, finance, and leadership on risk tolerance.

Key Points to Mention

  • Statistical significance: CI includes zero, so cannot conclude a positive effect.
  • Practical significance: Is a 3.5% lift worth the risk given refund increase?
  • Guardrail metrics: Refund rate increase is a red flag; investigate root cause.
  • Decision thresholds: Pre-defined thresholds for launch (e.g., CI entirely above 0, guardrails within acceptable range).
  • Extended test: Need more data to reduce uncertainty and monitor refunds.
  • Stakeholder communication: Tailor message to audience, focus on business impact and next steps.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.