← Expedia Interview Insights

Expedia·Data Scientist·Technical Phone Screen·Senior

Senior
Apr 2026

Summary

Expedia DS interview for their Search team, focused entirely on a hotel-ranking model case. The whole thing was one long multi-part question that spiraled through KPIs, experimentation, attribution, and profit estimation. Pretty intense, no fluff.

Questions Asked (7)

Q1

Define a single primary KPI for a hotel-ranking model serving enterprise clients who care about profitable bookings, and list at least three guardrail metrics with specific thresholds.

Product Analytics & MetricsProduct Strategy
Author's notes

I went with something like margin-adjusted bookings per search over a 7-day window, which felt right for the client value angle.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the business objective: enterprise clients care about profitable bookings, so the primary KPI should directly measure incremental profit from bookings influenced by the ranking model. Then propose guardrail metrics that ensure the model doesn't harm other critical aspects like customer experience, supplier relationships, or long-term value, with specific thresholds based on historical baselines or business rules.

Pro tip: Frame the KPI as a counterfactual or incremental metric (e.g., incremental profit per session) to isolate the model's impact, and set guardrail thresholds using A/B test power analysis or historical variance to avoid arbitrary numbers.

1. Clarify business goal and constraints

Confirm that the primary objective is maximizing profitable bookings for enterprise clients, and identify any constraints like customer satisfaction or supplier fairness.

2. Define the primary KPI

Propose a single primary KPI that directly measures incremental profit from bookings influenced by the ranking model, such as 'incremental profit per search session' or 'profit-weighted booking conversion rate'.

3. Identify potential negative side effects

Brainstorm metrics that could degrade if the model over-optimizes for profit, such as customer satisfaction, cancellation rates, or diversity of offerings.

4. Select guardrail metrics and set thresholds

Choose 3-4 guardrail metrics and assign specific thresholds (e.g., no more than 2% decrease in NPS, cancellation rate increase ≤ 0.5 percentage points) based on historical data or business tolerance.

5. Validate and iterate

Explain how you would monitor these metrics in A/B tests, and adjust thresholds if needed to balance profit and guardrails.

Key Points to Mention

  • Primary KPI should be profit-based, not just revenue or bookings, to align with enterprise clients' focus on profitability.
  • Use incremental metrics (e.g., lift in profit) to attribute impact to the ranking model.
  • Guardrail metrics should cover customer experience (e.g., NPS, CSAT), operational health (e.g., cancellation rate, customer service contacts), and supplier fairness (e.g., diversity of hotels shown).
  • Set thresholds based on historical baselines, statistical significance, or business rules (e.g., no more than 1% degradation).
  • Consider long-term metrics like repeat booking rate or customer lifetime value as guardrails.
  • Mention the importance of segmenting by enterprise client to ensure the model works across different clients.

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

Q2

How would you validate that offline proxy metrics like margin-weighted NDCG correlate with your online KPI? Describe the analysis design, acceptable correlation ranges, and what you'd do if they're misaligned.

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

This is where I got a little lost.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the validation as a meta-analysis across many historical A/B tests, where you compare the effect sizes on the offline proxy metric to the effect sizes on the online KPI. Then describe how you'd quantify the correlation (e.g., Spearman or Pearson) and set acceptable thresholds based on business risk. Finally, outline a decision framework for when correlation is weak, including metric refinement, segment analysis, or switching to a different proxy.

Pro tip: Emphasize that correlation should be assessed on the *relative* change (lift) between treatment and control, not absolute values, and that you need enough historical experiments with sufficient power to avoid noise. Also mention that a high correlation doesn't guarantee causality—you still need to validate on holdout experiments.

1. Collect historical experiment data

Gather a large set of past A/B tests where both the offline proxy metric and the online KPI were measured. Ensure each test has a clear treatment effect and sufficient sample size to detect it.

2. Compute effect sizes and correlation

For each experiment, calculate the relative lift (or standardized effect size) for both the proxy and the KPI. Then compute the correlation (e.g., Spearman rank) between these lifts across experiments.

3. Define acceptable correlation ranges

Set thresholds based on business impact: e.g., >0.7 strong, 0.4–0.7 moderate (use with caution), <0.4 weak (proxy unreliable). Consider confidence intervals and statistical significance of the correlation.

4. Diagnose misalignment

If correlation is weak, investigate causes: proxy not capturing key user behaviors, segment-specific effects, novelty effects, or metric definition issues. Use scatter plots and residual analysis to identify patterns.

5. Iterate and validate

Refine the proxy metric (e.g., reweight, add components) or consider alternative proxies. Re-run the correlation analysis on new experiments and validate with a holdout set before trusting the proxy for decision-making.

Key Points to Mention

  • Use of relative lift (treatment vs. control) rather than absolute metric values to control for experiment-specific baselines.
  • Choice of correlation metric: Spearman for robustness to non-linearity and outliers, Pearson for linear relationships.
  • Importance of statistical power and sample size in both offline and online measurements to avoid spurious correlations.
  • Segment-level analysis to check if correlation holds across key dimensions (e.g., device, market, user type).
  • Business context for acceptable correlation: higher stakes require stronger correlation; consider cost of false positives/negatives.
  • Actionable steps when misaligned: refine proxy, run additional experiments, or use proxy only for directional guidance with online validation.

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

Q3

Design an A/B test for this ranking model: randomization unit, traffic split, ramp strategy, MDE and sample size calculation, pre-experiment checks, and variance reduction techniques.

A/B Testing & Experimentation
Author's notes

Felt most comfortable here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the goal of the ranking model (e.g., increase bookings or revenue) and the unit of randomization (user or session). Then systematically walk through each component: randomization unit, traffic split, ramp strategy, MDE and sample size calculation, pre-experiment checks, and variance reduction techniques, ensuring you tie each choice back to business and statistical considerations.

Pro tip: Emphasize the importance of pre-experiment checks like A/A tests and sample ratio mismatch (SRM) to catch issues early, and mention that variance reduction techniques like CUPED can significantly speed up experiments without compromising validity.

1. Define Objective and Randomization Unit

Clarify the primary metric (e.g., conversion rate, revenue per user) and choose the randomization unit (e.g., user, session, or page) based on the model's application and potential interference.

2. Determine Traffic Split and Ramp Strategy

Decide on a 50/50 split for initial testing, then plan a gradual ramp-up (e.g., 1%, 5%, 10%, 50%) to monitor for technical issues and guardrail metrics before full rollout.

3. Calculate MDE and Sample Size

Based on business needs, set a minimum detectable effect (MDE) that is practically significant. Use power analysis (power=0.8, alpha=0.05) to compute required sample size, considering baseline variance and expected traffic.

4. Conduct Pre-Experiment Checks

Run A/A tests to validate the randomization and metric stability, check for sample ratio mismatch (SRM), and ensure no carryover effects from previous experiments.

5. Apply Variance Reduction Techniques

Use techniques like CUPED (Controlled-experiment Using Pre-Experiment Data) or stratification to reduce variance and increase sensitivity, enabling detection of smaller effects with the same sample size.

Key Points to Mention

  • Randomization unit: user-level for long-term effects, session-level for short-term; consider network effects if users interact.
  • Traffic split: start with 50/50 for balanced comparison, but consider unequal splits if risk is high.
  • Ramp strategy: gradual rollout to catch bugs and monitor guardrails (e.g., latency, error rates).
  • MDE and sample size: align MDE with business impact; use power analysis tools or formulas.
  • Pre-experiment checks: A/A tests, SRM checks, and metric stability over time.
  • Variance reduction: CUPED, stratification, or regression adjustment using pre-experiment covariates.

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

Q4

How would you detect and prevent metric gaming and novelty effects in this ranking experiment, and what would your rollback criterion be?

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

Novelty effects I handled okay, talked about monitoring engagement decay over time in the treatment group.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by outlining a comprehensive monitoring plan that tracks both primary and guardrail metrics, with novelty effects addressed through cohort analysis and extended observation periods. Then, detail specific detection methods like statistical tests for metric manipulation and novelty decay curves, and finally define a rollback criterion based on predefined thresholds for guardrail metrics and statistical significance.

Pro tip: Emphasize the importance of pre-registering your analysis plan and rollback criteria to avoid p-hacking and ensure objectivity. Also, mention that you would involve cross-functional partners (e.g., product, engineering) in defining what constitutes a 'gameable' metric and acceptable novelty periods.

1. Define Metrics and Guardrails

Identify primary success metrics (e.g., click-through rate, conversion) and guardrail metrics (e.g., user satisfaction, long-term retention) that are less susceptible to gaming. Establish clear definitions and data sources for each.

2. Detect Novelty Effects

Analyze metrics over time using cohort analysis to see if early gains diminish. Compare new users (who may show novelty) vs. existing users, and use holdout groups or extended A/B tests to measure long-term effects.

3. Detect Metric Gaming

Monitor for sudden, unnatural spikes in metrics, especially those easily manipulated (e.g., clicks). Use statistical process control, anomaly detection, and segment analysis to identify suspicious patterns. Cross-validate with qualitative feedback.

4. Prevent Gaming and Novelty Bias

Design experiments with multiple metrics, including guardrails, and use techniques like counterfactual analysis or synthetic control. Implement randomization checks and ensure balanced groups. Consider long-term holdout groups.

5. Establish Rollback Criteria

Predefine thresholds for guardrail metrics (e.g., if any guardrail degrades by >2% with 95% confidence) and for primary metrics if novelty is suspected. Include statistical significance and practical significance. Automate alerts for breaches.

Key Points to Mention

  • Use of guardrail metrics to detect negative side effects
  • Cohort analysis to separate novelty from true effect
  • Statistical methods like sequential testing or Bayesian approaches to monitor continuously
  • Pre-registration of analysis plan and rollback criteria to avoid bias
  • Cross-functional collaboration to define gaming and acceptable thresholds
  • Long-term holdout groups to measure sustained impact

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

Q5

Bookings are flat but cancellations rise 15% in the treatment group. Do you ship or rollback? Justify using expected profit impact and confidence intervals.

A/B Testing & ExperimentationPricing & Monetization
Author's notes

Rollback, pretty clearly.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying that the decision hinges on the net expected profit impact, not just the cancellation rate. Calculate the expected profit per user in both groups, incorporating booking volume, cancellation rate, and average profit per completed booking, then construct confidence intervals around the difference. Use the confidence interval to assess whether the treatment's profit impact is statistically and practically significant, and consider the business context (e.g., long-term customer value, seasonality) before making a ship/rollback recommendation.

Pro tip: Don't just focus on the cancellation rate; quantify the profit impact and its uncertainty. Also, consider segmenting the analysis to see if the effect is driven by a particular user group, which could inform a targeted rollout instead of a binary ship/rollback decision.

1. Define the profit metric

Establish the expected profit per user as the primary metric, incorporating booking volume, cancellation rate, and profit per completed booking. Clarify any assumptions about costs and revenues.

2. Calculate expected profit and difference

Compute the expected profit per user for both control and treatment groups. Calculate the difference (treatment minus control) to see the net impact.

3. Construct confidence intervals

Use statistical methods (e.g., bootstrapping or delta method) to construct a confidence interval for the difference in expected profit. Assess whether the interval includes zero and the width of the interval.

4. Interpret and decide

If the confidence interval is entirely above zero, ship; if entirely below zero, rollback. If it includes zero, consider the cost of uncertainty and potential long-term effects before deciding.

5. Consider business context and next steps

Evaluate whether the effect is consistent across segments, the potential for long-term impact, and whether a targeted rollout or further testing is warranted.

Key Points to Mention

  • Expected profit per user as the key metric, not just cancellation rate
  • Confidence intervals to quantify uncertainty in the profit difference
  • Statistical significance vs. practical significance
  • Potential segment-level analysis to identify heterogeneous treatment effects
  • Long-term customer value and repeat behavior
  • Business risk tolerance and cost of rolling out a potentially harmful change

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

Q6

Estimate the incremental gross profit per search with a 95% confidence interval, and explain how you'd separate true incrementality from attribution.

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

Blanked for a second on the incrementality framing.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the metric definition and the experimental design (e.g., randomized controlled trial) to estimate incremental gross profit per search. Then outline a statistical method to compute the point estimate and 95% confidence interval, and discuss how to isolate true incrementality from attribution biases using causal inference techniques.

Pro tip: Emphasize that incrementality requires a counterfactual—randomized experiments or quasi-experimental methods—and that attribution models alone cannot establish causality. Mention that you'd validate with holdout groups and sensitivity analyses to rule out confounding.

1. Define the metric and experiment

Clarify what 'incremental gross profit per search' means: profit from searches that would not have occurred without the intervention. Specify the randomization unit (user, session, search) and ensure the experiment is a controlled A/B test.

2. Estimate the point estimate and confidence interval

Compute the difference in mean gross profit per search between treatment and control groups. Use a two-sample t-test or bootstrap to derive the 95% confidence interval, accounting for clustering if needed.

3. Separate incrementality from attribution

Explain that attribution assigns credit to touchpoints but doesn't prove causality. Use randomized experiments (or quasi-experimental methods like geo-lift or synthetic control) to measure the true incremental effect by comparing treated vs. counterfactual outcomes.

4. Validate and address biases

Check for common biases: selection bias, spillover effects, and novelty effects. Use holdout groups, pre-period trend analysis, and sensitivity checks to ensure the incremental estimate is robust.

Key Points to Mention

  • Randomized controlled trials (A/B tests) as the gold standard for measuring incrementality.
  • Difference-in-differences or synthetic control methods when randomization isn't possible.
  • The importance of defining the counterfactual: what would have happened without the intervention?
  • Confidence intervals: use bootstrapping or t-tests, and consider clustering by user or session.
  • Attribution models (last-click, multi-touch) are descriptive, not causal—they can overstate incrementality.
  • Holdout groups and long-term holdouts to measure sustained incremental impact.

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

Q7

For paid channels, propose a multi-touch attribution approach and explain how you'd calibrate it against holdout or geo-based tests.

Product Analytics & MetricsA/B Testing & ExperimentationPricing & Monetization
Author's notes

I went with a Shapley value approach for the multi-touch piece since it handles credit allocation in a theoretically clean way, then mentioned geo holdout experiments as the calibration layer.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by outlining a multi-touch attribution model that balances simplicity and accuracy, such as a data-driven Markov chain or Shapley value approach, and explain how it assigns credit across channels. Then describe a calibration plan using holdout or geo-based tests to validate and adjust the model, emphasizing incremental lift measurement and bias correction.

Pro tip: Emphasize that attribution models are directional and should be calibrated with experiments to measure true incrementality; mention that you'd set up a continuous feedback loop to update the model as channel dynamics change.

1. Choose a Multi-Touch Attribution Model

Select a model like Markov chain or Shapley value that captures channel interactions and assigns fractional credit. Justify the choice based on data availability, business goals, and interpretability.

2. Define Calibration Metrics and Design Tests

Identify key metrics (e.g., incremental conversions, ROAS) and design holdout or geo-based tests to measure true causal impact. Ensure test groups are statistically comparable and account for spillover effects.

3. Run Experiments and Collect Data

Execute the tests, ensuring proper randomization and sufficient power. Collect data on conversions and spend for both test and control groups, and monitor for external validity threats.

4. Compare and Calibrate Model Outputs

Compare attribution model predictions with experimental lift to identify biases. Adjust model parameters or apply a calibration factor to align with incremental results.

5. Implement Continuous Monitoring and Iteration

Set up ongoing validation to detect drift and recalibrate as needed. Use a feedback loop to refine the model with new experimental data.

Key Points to Mention

  • Data-driven attribution models (e.g., Markov chain, Shapley value) and their trade-offs
  • Holdout tests (user-level or geo-based) to measure incrementality
  • Geo-based tests: matched market pairs, synthetic control, or switchback designs
  • Calibration techniques: bias correction, scaling factors, or Bayesian updating
  • Challenges: spillover effects, selection bias, and time-lag in conversions
  • Integration with marketing mix modeling (MMM) for a holistic view

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