← Netflix Interview Insights

Netflix·Data Scientist·Technical Phone Screen·Senior

SeniorPrefer not to say
Jun 2026

Summary

Interviewed at Netflix, got hit with a stats question that I was not fully prepared for. Short and focused, but it stuck with me.

Questions Asked (1)

Q1

What do you do when your test metric is not normally distributed?

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

I fumbled around with 'use a non-parametric test' and mentioned the Mann-Whitney U, which was fine but felt incomplete.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying that non-normality is common in metrics like revenue or session duration, and it doesn't automatically invalidate the test. Then walk through a structured diagnostic and remediation process, emphasizing that the right fix depends on sample size, metric type, and business context.

Pro tip: At Netflix scale, with millions of users, the Central Limit Theorem often makes the t-test robust to non-normality; but for heavy-tailed metrics, consider capping outliers or using a bootstrap to get reliable confidence intervals.

1. Diagnose the distribution

Plot the metric distribution and compute skewness/kurtosis to understand the nature and severity of non-normality. Check if it's due to outliers, zero-inflation, or a naturally skewed metric like revenue.

2. Assess impact on the test

Determine if the sample size is large enough for the Central Limit Theorem to apply, or if the non-normality is severe enough to bias results. Consider the test's sensitivity and the metric's importance.

3. Choose a remediation strategy

Select an appropriate method: transform the metric (e.g., log), use non-parametric tests (e.g., Mann-Whitney), bootstrap the sampling distribution, or apply robust methods like trimmed means. For heavy tails, consider capping outliers.

4. Validate and communicate

Run simulations or sensitivity analyses to confirm the chosen method's validity. Clearly communicate the assumptions and limitations to stakeholders, and align on the decision-making framework.

Key Points to Mention

  • Central Limit Theorem and sample size considerations
  • Common non-normal metrics in tech: revenue, session time, clicks per user
  • Transformations (log, Box-Cox) and their interpretability trade-offs
  • Non-parametric tests (Mann-Whitney, bootstrap) and when to use them
  • Outlier handling: capping, winsorizing, or trimming
  • Netflix-specific context: large scale, business impact of metrics, and experimentation culture

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