← PayPal Interview Insights

PayPal·Data Scientist·Take-home Assignment·Senior

Senior
Jul 2026

Summary

PayPal DS take-home before onsite: you get a dataset, analyze an A/B test, and present slides. The scope was wider than I expected and the noncompliance angle tripped me up a bit.

Questions Asked (5)

Q1

What primary metric and guardrail metrics would you use to decide whether to ship the experiment?

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

I went with orders_7d as primary and flagged revenue_7d and support_tickets_7d as guardrails.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the business context, then define one primary metric that directly measures success and 2-3 guardrail metrics that ensure no harm to user experience or business health. Explain how you would set thresholds, monitor during the test, and make a ship/no-ship decision based on statistical significance and practical significance.

Pro tip: Always tie your metrics to the company's north-star metric and mention that you would pre-register your primary metric and guardrails to avoid p-hacking. Also, consider segment-level analysis to catch heterogeneous treatment effects that might be masked in the overall average.

1. Clarify experiment goal and hypothesis

Ask about the specific change being tested and the expected impact. Align on the business objective (e.g., increase revenue, engagement) to select relevant metrics.

2. Choose one primary metric

Select a single metric that directly measures the desired outcome and is sensitive to the change. It should be tied to the business goal and have a clear direction (e.g., conversion rate).

3. Define guardrail metrics

Identify 2-3 metrics that ensure the change doesn't harm other critical areas, such as user experience (e.g., page load time), revenue (e.g., average order value), or long-term health (e.g., retention).

4. Set decision criteria and thresholds

Determine minimum detectable effect, statistical significance level, and practical significance. Define what constitutes a win, loss, or inconclusive result for both primary and guardrail metrics.

5. Analyze results and make recommendation

Check if primary metric improved significantly without guardrails degrading beyond acceptable thresholds. Consider segment analysis and qualitative insights before shipping.

Key Points to Mention

  • Primary metric should be directly tied to the experiment's hypothesis and business goal, e.g., conversion rate for a checkout flow change.
  • Guardrail metrics protect against unintended consequences, e.g., latency, error rates, customer support contacts, or revenue per user.
  • Statistical significance (p-value < 0.05) and practical significance (effect size) both matter for decision-making.
  • Pre-register metrics and thresholds to avoid p-hacking and ensure valid inference.
  • Consider novelty effects and long-term impact by running the experiment for sufficient duration.
  • Segment analysis can reveal heterogeneous treatment effects that might be hidden in aggregate results.

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

Q2

Walk through the statistical tests and confidence intervals you would run, and what sanity checks you would perform before trusting any results.

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

SRM check first, always.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around the experiment lifecycle: design, execution, analysis, and validation. Emphasize that you choose statistical tests based on the metric type and experiment design, and that you always perform sanity checks before interpreting results. Highlight the importance of practical significance alongside statistical significance, especially in a fintech context like PayPal.

Pro tip: Mention that you pre-register your analysis plan and guard against peeking, and that you check for novelty effects and seasonality—these show maturity beyond textbook stats.

1. Clarify the Experiment Design and Metric

Identify the primary metric (e.g., conversion rate, revenue per user), its distribution, and the randomization unit. Confirm the hypothesis, sample size, and power analysis.

2. Select Appropriate Statistical Tests

For continuous metrics, use t-tests or Mann-Whitney U if assumptions fail; for binary metrics, use proportion tests or chi-square. Consider sequential testing or Bayesian methods if needed.

3. Compute Confidence Intervals

Calculate CIs for the difference in means or proportions, using methods like bootstrap or delta method for ratio metrics. Report both absolute and relative lift.

4. Run Sanity Checks

Verify sample ratio mismatch (SRM), check for outliers, ensure no data leakage, and validate that pre-experiment covariates are balanced. Also check for novelty effects and segment-level anomalies.

5. Interpret Results with Business Context

Assess practical significance, consider multiple testing corrections, and evaluate impact on key business metrics. Recommend next steps based on robustness of findings.

Key Points to Mention

  • Sample Ratio Mismatch (SRM) check to ensure randomization integrity
  • Choice of test based on metric type and distribution (e.g., t-test vs. Mann-Whitney)
  • Confidence intervals and effect sizes, not just p-values
  • Multiple testing correction (e.g., Bonferroni, Benjamini-Hochberg) when analyzing multiple metrics
  • Novelty and primacy effects, especially in user behavior experiments
  • Practical significance and business impact, not just statistical significance

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

Q3

How would you handle the fact that some assigned users never actually saw the new experience, and which estimand would you report?

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

This is where I fumbled a bit.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, clarify the distinction between intent-to-treat (ITT) and treatment-on-the-treated (TOT) estimands, and explain how to diagnose non-compliance by measuring actual exposure. Then, recommend reporting ITT as the primary estimand for its policy relevance and unbiasedness, while using TOT (via instrumental variables or CACE) as a secondary analysis to estimate the effect among those who actually saw the new experience. Finally, discuss practical steps to mitigate non-compliance in future experiments, such as improving logging or user targeting.

Pro tip: Emphasize that ITT preserves randomization and reflects real-world impact, which is often what stakeholders care about; but also show you understand TOT's value for understanding the mechanism. Mention that you would pre-register the primary estimand to avoid post-hoc cherry-picking.

1. Diagnose non-compliance

Quantify the proportion of assigned users who never saw the new experience and investigate reasons (e.g., technical issues, user behavior). Check if non-compliance is differential across treatment and control groups.

2. Define estimands

Clearly define ITT (effect of assignment) and TOT (effect of actual exposure). Explain that ITT is unbiased and policy-relevant, while TOT estimates the effect for compliers but requires stronger assumptions.

3. Choose primary estimand

Recommend ITT as the primary estimand because it preserves randomization and answers 'what happens if we roll out this feature?' It is also robust to non-compliance.

4. Conduct secondary analysis

If needed, estimate TOT using instrumental variables (e.g., two-stage least squares) or CACE, but clearly state assumptions (e.g., exclusion restriction, monotonicity) and interpret with caution.

5. Recommend improvements

Suggest ways to reduce non-compliance in future experiments, such as better logging, user education, or ensuring feature visibility, and discuss how to handle it in analysis (e.g., per-protocol analysis as sensitivity check).

Key Points to Mention

  • Intent-to-treat (ITT) vs. treatment-on-the-treated (TOT) estimands
  • Non-compliance and its impact on causal inference
  • Instrumental variables (IV) or complier average causal effect (CACE) for TOT
  • Assumptions for TOT: exclusion restriction, monotonicity, and independence
  • Policy relevance of ITT for business decisions
  • Pre-registration of primary estimand to avoid bias

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

Q4

How would you investigate heterogeneous treatment effects across platform, country, and user type without running into p-hacking issues?

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

Pre-registration is the real answer here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the problem as a multiple comparisons challenge and propose a pre-registered analysis plan with a hierarchy of hypotheses. Then describe using methods like Bonferroni or Benjamini-Hochberg to control error rates, and emphasize the importance of validation through holdout or replication. Finally, discuss practical trade-offs between statistical rigor and actionable insights.

Pro tip: Mention that you would pre-specify the subgroups and interactions of interest before looking at the data, and use a holdout set or sequential testing to confirm findings. This shows you understand that p-hacking often arises from post-hoc fishing.

1. Pre-register hypotheses and subgroups

Define the primary metric, the subgroups (platform, country, user type), and the specific interactions you will test before analyzing the data. This prevents data dredging and ensures confirmatory analysis.

2. Choose appropriate multiple testing correction

Select a method like Bonferroni, Holm-Bonferroni, or Benjamini-Hochberg to control family-wise error rate or false discovery rate, depending on the number of tests and the cost of false positives.

3. Use hierarchical or Bayesian methods

Consider hierarchical models (e.g., partial pooling) or Bayesian approaches with informative priors to borrow strength across subgroups and reduce variance, which can mitigate multiple comparisons issues.

4. Validate findings on holdout or replication

Split the data or run a follow-up experiment to confirm any significant heterogeneous treatment effects, ensuring results are not spurious.

5. Communicate uncertainty and practical significance

Report effect sizes with confidence intervals, and discuss whether the detected heterogeneity is large enough to inform business decisions, balancing statistical and practical significance.

Key Points to Mention

  • Multiple comparisons problem and family-wise error rate (FWER) vs false discovery rate (FDR)
  • Pre-registration and confirmatory vs exploratory analysis
  • Bonferroni, Holm-Bonferroni, Benjamini-Hochberg procedures
  • Hierarchical models / partial pooling / Bayesian shrinkage
  • Holdout validation and replication
  • Power analysis and sample size considerations for subgroup analyses

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

Q5

What would your slide deck look like, and how would you communicate uncertainty and next steps to a PM or engineering audience?

Cross-functional AlignmentA/B Testing & Experimentation
Author's notes

Honestly the communication part felt like the real test.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Structure your answer around a clear narrative: start with the objective and key findings, then present the evidence with appropriate uncertainty quantification, and end with actionable next steps tailored to the audience. Emphasize how you would adapt the communication for PMs (focus on business impact and decisions) versus engineers (focus on methodology and implementation details).

Pro tip: Use a 'confidence ladder' to communicate uncertainty: label findings as 'high confidence' (statistically significant, robust), 'medium confidence' (directional, needs validation), or 'low confidence' (exploratory, hypothesis-generating). This helps stakeholders calibrate their decisions without drowning in statistical jargon.

1. Set the context and objective

Briefly remind the audience of the business question or hypothesis being tested, and why it matters. This aligns everyone before diving into details.

2. Present key findings with uncertainty

Show the main results using clear visuals (e.g., confidence intervals, effect sizes) and explicitly state the level of uncertainty. Avoid overclaiming; use phrases like 'we observed a X% lift with 95% confidence' or 'the data suggests a trend, but we need more data to confirm.'

3. Explain methodology and limitations

For engineers, include details on experimental design, sample size, and potential biases. For PMs, summarize in plain language: what we did, what we can and cannot conclude, and any caveats.

4. Propose next steps and recommendations

Offer clear, prioritized actions based on the findings. For PMs, tie to business impact (e.g., 'recommend rolling out to 10% of users to monitor long-term effects'). For engineers, suggest technical follow-ups (e.g., 'run a power analysis for a follow-up test').

5. Invite discussion and address questions

Open the floor for questions, and be prepared to dive deeper into any area. This shows confidence and collaborative spirit.

Key Points to Mention

  • Use of confidence intervals and p-values to quantify uncertainty
  • Distinction between statistical significance and practical significance
  • Tailoring communication: business metrics for PMs, technical details for engineers
  • Acknowledging limitations and potential biases in the analysis
  • Proposing actionable next steps with clear ownership and timelines
  • Using visualizations that effectively convey uncertainty (e.g., error bars, Bayesian posterior distributions)

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