← Airbnb Interview Insights

Airbnb·Data Scientist·Technical Phone Screen·Senior

Senior
May 2026

Summary

Airbnb data scientist interview that went deep into metric design and experimentation rigor. The question was essentially one long technical prompt covering everything from logging integrity to power calculations, and it felt less like a conversation and more like defending a thesis.

Questions Asked (5)

Q1

For a newly launched feature, define a primary success metric tied to long-term value. Walk through the exact formula including numerator/denominator definitions, deduplication logic, windowing choices, and how you handle outliers and bot traffic.

Product Analytics & MetricsA/B Testing & Experimentation
Author's notes

This is where I spent most of my time and also where I stumbled first.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the feature's goal and the long-term value it aims to create, then propose a primary metric that captures that value (e.g., repeat bookings or host retention). Walk through the metric's formula with precise numerator/denominator definitions, deduplication, windowing, and outlier/bot handling, emphasizing how each choice aligns with long-term impact.

Pro tip: Tie the metric to a north-star business outcome and explicitly discuss trade-offs (e.g., short-term engagement vs. long-term retention) to show strategic thinking. Also, mention that you would validate the metric through A/B testing and monitor for unintended consequences.

1. Clarify feature and long-term value

Ask clarifying questions to understand the feature's purpose and the long-term value it should drive (e.g., increased trust, repeat usage). Align the metric with a strategic business goal.

2. Define the primary metric

Propose a metric that directly measures the long-term value, such as 'repeat booking rate within 90 days' or 'host retention rate after 6 months'. Explain why it's the best proxy for long-term success.

3. Specify formula and definitions

Detail the numerator and denominator: e.g., numerator = number of unique users who made a repeat booking within the window; denominator = number of unique users exposed to the feature. Define deduplication logic (e.g., count each user once) and windowing choices (e.g., 90-day window from first exposure).

4. Handle outliers and bot traffic

Describe how to identify and exclude bot traffic (e.g., using user-agent, behavior patterns) and handle outliers (e.g., winsorizing, capping extreme values, or using robust statistics). Explain the impact on the metric.

5. Validate and iterate

Mention that you would validate the metric via A/B testing, check for guardrail metrics, and iterate if the metric doesn't correlate with long-term value. Discuss potential pitfalls and how to address them.

Key Points to Mention

  • Alignment with long-term business value (e.g., retention, repeat usage)
  • Precise numerator and denominator definitions with clear inclusion/exclusion criteria
  • Deduplication logic to avoid double-counting users or events
  • Windowing choices (e.g., 30/60/90 days) and rationale for the window length
  • Outlier handling techniques (e.g., winsorizing, capping) and their impact
  • Bot traffic detection and exclusion methods (e.g., user-agent filtering, behavioral analysis)

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

Q2

What are the trade-offs between ratio-of-sums and sum-of-ratios when computing your primary metric, and how does your choice affect sensitivity to late or backfilled events?

A/B Testing & ExperimentationTechnical Trade-offsProduct Analytics & Metrics
Author's notes

Blanked for a second on the late events angle.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining both metrics and their typical use cases, then discuss how each handles late or backfilled events in terms of sensitivity and bias. Use a concrete example (e.g., click-through rate) to illustrate the trade-offs, and conclude with a recommendation based on the specific context and goals.

Pro tip: Mention that ratio-of-sums is generally more robust to late events because it aggregates totals before taking the ratio, while sum-of-ratios can be more sensitive to outliers and late data, but may be preferred when you need to weight each unit equally. Also, highlight the importance of aligning the metric with the randomization unit and the analysis unit.

1. Define the metrics

Clearly define ratio-of-sums (e.g., total clicks / total views) and sum-of-ratios (e.g., average of per-user click-through rates). Explain that the choice depends on the unit of analysis and the goal of the experiment.

2. Discuss trade-offs

Compare the two approaches in terms of interpretability, statistical power, and sensitivity to outliers. Ratio-of-sums is typically more stable and easier to interpret, while sum-of-ratios can be dominated by users with few events.

3. Analyze sensitivity to late/backfilled events

Explain that late events affect both numerator and denominator, but ratio-of-sums may be less sensitive because it aggregates across all users. Sum-of-ratios can be more sensitive if late events disproportionately affect users with extreme ratios.

4. Provide a concrete example

Use a scenario like measuring click-through rate: if a user clicks late, their ratio changes, but in ratio-of-sums the overall ratio may not shift much if many users are included. In sum-of-ratios, that user's ratio could have a large impact if they have few views.

5. Recommend based on context

Conclude that the choice depends on the metric's purpose, the randomization unit, and the expected delay patterns. Suggest using ratio-of-sums for global metrics and sum-of-ratios when per-unit effects are of interest, and always monitor for late data.

Key Points to Mention

  • Ratio-of-sums is equivalent to a weighted average of per-unit ratios, weighted by the denominator.
  • Sum-of-ratios gives equal weight to each unit, which can be desirable if units are the randomization unit.
  • Late events can introduce bias if they are correlated with treatment, so it's crucial to wait for data maturity or use statistical corrections.
  • Backfilled events can affect both metrics, but ratio-of-sums may be more robust because it pools data across units.
  • Consider using a delta method or bootstrap for variance estimation when using sum-of-ratios.
  • Always align the metric with the experiment's goal and the unit of randomization to avoid misinterpretation.

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

Q3

How would you detect and correct a silent logging change that occurs mid-experiment, and what does timezone misalignment do to your metric computation?

Root Cause AnalysisA/B Testing & ExperimentationData Modeling
Author's notes

The 'silent' part is the key word here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining how you would detect a silent logging change through automated data quality checks and anomaly detection on log volumes and metric distributions. Then describe a correction strategy that involves isolating the affected period, adjusting metrics using pre-change baselines, and validating with a holdout. Finally, discuss how timezone misalignment can shift daily boundaries, causing misattribution of events and biased experiment results.

Pro tip: Emphasize the importance of logging schema versioning and timezone-aware timestamps in experiment design to prevent such issues. Mention that at Airbnb, where users span multiple timezones, aligning on UTC for analysis but local time for user experience is critical.

1. Detection

Implement automated data quality checks (e.g., row counts, null rates, distribution shifts) and anomaly detection to flag sudden changes in logging behavior. Monitor log volume and key metric trends in real-time.

2. Diagnosis

Investigate the root cause by comparing log schemas, timestamps, and event definitions before and after the change. Check deployment logs and code changes to identify when and why the logging change occurred.

3. Correction

Isolate the affected period and adjust metrics by re-weighting or using pre-change baselines. If possible, re-run the experiment or use a holdout group to validate corrections.

4. Timezone Impact

Explain how timezone misalignment shifts daily boundaries, causing events to be assigned to wrong days, leading to biased daily metrics and incorrect experiment conclusions. Use UTC for analysis and ensure all timestamps are timezone-aware.

5. Prevention

Advocate for logging schema versioning, timezone standardization, and automated monitoring to prevent future silent changes. Include data quality checks in the experiment pipeline.

Key Points to Mention

  • Automated data quality checks and anomaly detection for logging changes
  • Log schema versioning and change management
  • Timezone-aware timestamps and UTC normalization for analysis
  • Impact of timezone misalignment on daily metrics and experiment validity
  • Correction methods: re-weighting, baseline adjustment, holdout validation
  • Root cause analysis using deployment logs and code diffs

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

Q4

Given historical variance from past experiments, compute the minimum detectable effect and required sample size for a 14-day test. Then explain how you'd approach sequential monitoring and alpha spending across that window.

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

Power calculations I can do in my sleep so this part was fine.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying assumptions and defining the metric, then walk through the MDE and sample size calculation using historical variance, power, and significance level. Finally, discuss sequential monitoring with alpha spending functions to control false positives over the 14-day window.

Pro tip: Emphasize that sequential monitoring requires pre-registration of the spending function and that peeking without adjustment inflates Type I error; mention Airbnb's likely use of tools like sequential testing or group sequential designs.

1. Clarify assumptions and inputs

Confirm the metric (e.g., conversion rate), baseline variance from historical experiments, desired power (typically 80%), significance level (5%), and whether the test is one- or two-sided.

2. Compute MDE and sample size

Use the formula for MDE given sample size or solve for sample size given MDE, incorporating historical variance. For a 14-day test, consider daily traffic to ensure feasibility.

3. Explain sequential monitoring

Describe how to monitor results daily while controlling Type I error. Introduce alpha spending functions (e.g., O'Brien-Fleming, Pocock) that allocate alpha over the 14 days.

4. Discuss trade-offs and practical considerations

Address trade-offs between early stopping and power, the impact of multiple looks, and how to choose a spending function based on business needs (e.g., early detection vs. strict error control).

Key Points to Mention

  • MDE formula: MDE = (Z_{α/2} + Z_β) * sqrt(2 * σ^2 / n) for equal groups, where σ^2 is variance.
  • Sample size calculation: n = (2 * (Z_{α/2} + Z_β)^2 * σ^2) / MDE^2 per group.
  • Sequential monitoring requires alpha spending to control overall Type I error rate.
  • Common alpha spending functions: O'Brien-Fleming (conservative early, liberal late), Pocock (constant alpha at each look).
  • Pre-registration of monitoring plan and spending function is critical to avoid p-hacking.
  • Consider using sequential testing methods like always-valid p-values or group sequential designs.

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

Q5

Define at least three guardrail metrics for the experiment, set canary thresholds for each, and explain what you'd do if the primary metric improves but one guardrail shows a slight regression.

A/B Testing & ExperimentationProduct Analytics & MetricsTechnical Trade-offs
Author's notes

Picked crash rate, p95 latency, and complaint rate.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining three guardrail metrics that cover different risk areas (e.g., user experience, business health, system performance) and set canary thresholds based on historical variability and practical significance. Then, outline a decision framework for when the primary metric improves but a guardrail regresses slightly, emphasizing investigation, risk assessment, and potential mitigation before deciding to ship or iterate.

Pro tip: Frame guardrails as 'canary metrics' that act as early warning signals—set thresholds at a level that balances sensitivity and false positives, and always consider the cost of a false negative (shipping a harmful change) versus a false positive (delaying a beneficial change).

1. Select guardrail metrics

Choose three metrics that capture potential negative impacts: e.g., user engagement (e.g., booking completion rate), business health (e.g., revenue per user), and system performance (e.g., page load time). Ensure they are sensitive to changes and aligned with company goals.

2. Set canary thresholds

For each guardrail, define a threshold (e.g., -2% relative change) based on historical variance, minimum detectable effect, and business tolerance. Use statistical significance and practical significance to avoid overreacting to noise.

3. Monitor and detect regression

During the experiment, continuously monitor guardrails. If a guardrail shows a slight regression (e.g., -1.5% when threshold is -2%), flag it for investigation but do not automatically stop the experiment.

4. Investigate and assess trade-offs

If primary metric improves but a guardrail regresses slightly, dig into segment-level data, check for novelty effects, and quantify the net impact. Consider whether the regression is acceptable given the primary gain.

5. Decide and act

Based on investigation, either ship with monitoring, iterate to mitigate the guardrail regression, or abandon if the risk outweighs benefits. Document the decision and learnings.

Key Points to Mention

  • Guardrail metrics should cover different dimensions: user experience, business metrics, and technical performance.
  • Canary thresholds should be set using statistical power analysis and historical data, not arbitrary numbers.
  • A slight regression in a guardrail doesn't automatically mean failure; consider the magnitude, confidence interval, and business context.
  • Investigate root causes: check for segment-specific effects, novelty, or interactions with other changes.
  • Quantify the trade-off: calculate the net impact (e.g., primary metric gain vs. guardrail loss) in terms of business value.
  • Consider mitigation strategies: can you adjust the feature to reduce the guardrail regression while preserving the primary gain?

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