← Uber Interview Insights

Uber·Data Scientist·Technical Phone Screen·Senior

SeniorPrefer not to say
Jun 2026

Summary

Case-style interview for a data scientist role at Uber, all three questions were marketplace and experimentation focused. No fluff, just dense product-analytics scenarios that assumed you already knew the basics and wanted to see how far you could push.

Questions Asked (3)

Q1

The company wants to add an in-app warning that tells drivers when a passenger's drop-off location is in a high-risk neighborhood. How would you design an experiment to evaluate this feature, and what metrics would you track across business, safety, marketplace, and fairness dimensions? How does running the test across multiple cities change your approach?

A/B Testing & ExperimentationProduct Analytics & MetricsProduct Sense & Ideation
Author's notes

This one is harder than it looks because the fairness angle is a trap you can walk into without realizing.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the feature's goal and defining a clear hypothesis, then outline a randomized controlled experiment with appropriate randomization unit and guardrail metrics. Structure your answer around the four metric dimensions (business, safety, marketplace, fairness) and discuss how multi-city testing introduces heterogeneity that requires stratified randomization, city-level fixed effects, and careful analysis of treatment effect variation.

Pro tip: Acknowledge that this feature could backfire by stigmatizing neighborhoods or causing drivers to avoid certain areas, so include fairness and marketplace health as guardrails, not just afterthoughts. Also, propose a pre-registered analysis plan to avoid p-hacking across cities.

1. Clarify goal and hypothesis

Define the feature's intended effect (e.g., increase driver safety perception) and state a testable hypothesis. Identify the primary decision metric and minimum detectable effect.

2. Design the experiment

Choose randomization unit (driver, trip, or city), treatment/control groups, and duration. Consider switchback or cluster randomization if spillover is a concern. Pre-register the analysis plan.

3. Select metrics across dimensions

Define business (e.g., completed trips, driver retention), safety (e.g., incidents, driver reports), marketplace (e.g., ETAs, acceptance rates, cancellations), and fairness (e.g., differential impact on low-income or minority neighborhoods) metrics.

4. Address multi-city testing

Stratify randomization by city, include city fixed effects, and test for treatment effect heterogeneity. Use hierarchical models or meta-analysis to pool results while accounting for city-level variance.

5. Analyze and interpret results

Check for statistical significance, practical significance, and guardrail violations. Investigate heterogeneous effects and potential unintended consequences before making a recommendation.

Key Points to Mention

  • Randomization unit and potential spillover effects (e.g., drivers in control group learning about the feature)
  • Guardrail metrics to detect harm, such as increased cancellations or reduced driver willingness to serve certain areas
  • Fairness metrics: measure differential impact on protected groups or neighborhoods, and consider whether the feature reinforces bias
  • Multi-city heterogeneity: use city-level fixed effects, stratified randomization, and test for interaction effects
  • Statistical power and minimum detectable effect, especially when splitting by city
  • Long-term vs short-term effects: consider novelty effects and whether to run a holdout group for longer-term measurement

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

Q2

Why does ETA accuracy matter so much in a ride-hailing product? What metrics would you use to measure it, and what approaches would you take to diagnose and improve it?

Product Analytics & MetricsRoot Cause AnalysisProduct Sense & Ideation
Author's notes

Felt like a warmup but they pushed hard on the ML side.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by linking ETA accuracy to core business metrics like rider trust, conversion, and driver utilization, then define a metric framework that captures both bias and precision. For diagnosis, segment errors by time, location, and trip characteristics, and propose improvements across data, modeling, and product interventions.

Pro tip: Emphasize that ETA accuracy is not just about average error but about the tails—large errors disproportionately damage trust and retention. Also, consider the cost of over- vs under-prediction: under-prediction frustrates riders, while over-prediction may reduce conversion.

1. Business Impact

Explain why ETA accuracy matters: it drives rider trust, conversion, retention, and driver efficiency. Inaccurate ETAs lead to cancellations, poor ratings, and lost revenue.

2. Metric Definition

Define metrics that capture both central tendency and tails: MAE, RMSE, bias (mean error), percentage of trips with error > X minutes, and correlation with actual wait times. Also track business metrics like cancellation rate and conversion.

3. Diagnosis

Segment errors by time of day, geography, trip distance, traffic conditions, and driver/rider behavior. Use residual analysis and error decomposition to identify systematic biases vs. random noise.

4. Improvement Strategies

Propose improvements: better data (real-time traffic, historical patterns), advanced modeling (ML with features like weather, events), and product changes (e.g., showing ranges, updating ETAs dynamically).

5. Validation & Monitoring

Set up A/B tests to measure impact of changes on both ETA accuracy and business metrics. Continuously monitor performance and iterate.

Key Points to Mention

  • ETA accuracy directly impacts rider trust, conversion, and retention; inaccurate ETAs lead to cancellations and poor ratings.
  • Use both accuracy metrics (MAE, RMSE) and bias metrics (mean error) to capture systematic over/under-prediction.
  • Focus on tail errors (e.g., 90th percentile error) because large errors disproportionately affect user experience.
  • Segment analysis by time, location, trip type, and traffic conditions to identify root causes.
  • Leverage real-time data (traffic, weather, events) and machine learning models to improve predictions.
  • Consider product interventions like showing ETA ranges or dynamic updates to manage expectations.

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

Q3

Uber wants to notify drivers a week in advance about a promotion they can earn by being online in specific locations during specific time windows. What effects could this create, what's wrong with a simple user-level A/B test here, what issues come up with a switchback design, and how would you estimate variance under a synthetic control approach?

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

Probably the hardest of the three.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by discussing the potential behavioral and marketplace effects of the promotion, then explain why a simple user-level A/B test is invalid due to interference and spillover. Next, analyze the switchback design's challenges, and finally outline how to estimate variance under a synthetic control approach, emphasizing the need for counterfactual construction and uncertainty quantification.

Pro tip: Demonstrate awareness that in marketplace experiments, the unit of randomization must align with the unit of interference; often this means randomizing at the market or time level, and using methods like switchback or synthetic control to handle spillovers.

1. Identify potential effects

Consider how the promotion could change driver behavior (e.g., increased online time, relocation to specific areas, gaming the system) and marketplace outcomes (e.g., supply-demand balance, rider wait times, prices). Also think about unintended consequences like cannibalization or inequitable distribution.

2. Explain why user-level A/B test fails

A simple user-level A/B test suffers from interference: treated drivers may affect control drivers through shared marketplace dynamics (e.g., competition for rides, surge pricing). This violates SUTVA and biases treatment effect estimates.

3. Analyze switchback design issues

Switchback designs alternate treatment across time periods for the same unit (e.g., city). Issues include carryover effects (drivers may anticipate or remember promotions), time-varying confounders (e.g., weather, events), and reduced power due to temporal correlation. Also, the treatment is at the time level, so individual driver effects are not directly estimable.

4. Estimate variance under synthetic control

Synthetic control constructs a counterfactual by weighting control units (e.g., other cities) to match the treated unit's pre-treatment characteristics. Variance estimation can be done via placebo tests (in-space and in-time), bootstrap, or by modeling the synthetic control's prediction error. Key is to account for uncertainty in the weights and the counterfactual.

Key Points to Mention

  • Interference and spillover effects in marketplace experiments
  • SUTVA violation and its consequences for A/B testing
  • Carryover effects and time-varying confounders in switchback designs
  • Synthetic control method: constructing counterfactual, placebo tests, and variance estimation
  • Importance of choosing the right randomization unit (e.g., city, time) to match interference structure
  • Potential behavioral responses like strategic timing or location choices by drivers

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