← Uber Interview Insights

Uber·Data Scientist·Technical Phone Screen·Senior

Senior
Jun 2026

Summary

Uber DS interview with a deep experiment design question about promo allocation. The whole session was basically one long case that kept branching into more and more technical sub-problems. Felt like a lot to hold in your head at once.

Questions Asked (6)

Q1

Design a controlled experiment to test whether a new model (M1) outperforms the existing model (M0) for targeting $5 promotional offers to users.

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

This was the core question and it sprawled in every direction.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the experiment's goal, primary metric, and guardrail metrics, then outline the randomization unit and sample size calculation. Describe the experimental design, including control and treatment groups, and how you would analyze the results to determine if M1 significantly outperforms M0. Finally, discuss potential pitfalls and how to mitigate them.

Pro tip: Emphasize the importance of choosing the right randomization unit (e.g., user or session) to avoid interference and ensure valid inference, and mention the need for a sufficiently long experiment to capture delayed effects like redemptions.

1. Define Hypothesis and Metrics

Clearly state the null and alternative hypotheses. Identify the primary metric (e.g., redemption rate) and guardrail metrics (e.g., user retention, revenue) to ensure the new model doesn't harm other aspects.

2. Design Experiment

Choose the randomization unit (e.g., user ID) and determine the sample size using power analysis. Randomly assign users to control (M0) and treatment (M1) groups, ensuring balance on key covariates.

3. Run Experiment and Collect Data

Launch the experiment, ensuring proper logging and data collection. Monitor for any technical issues or sample ratio mismatch (SRM) that could invalidate results.

4. Analyze Results

Compare the primary metric between groups using appropriate statistical tests (e.g., t-test or bootstrap). Check for statistical significance and practical significance, and analyze guardrail metrics.

5. Interpret and Decide

Conclude whether M1 outperforms M0 based on the primary metric without harming guardrails. Consider segment-level analyses and potential long-term effects before making a recommendation.

Key Points to Mention

  • Randomization unit: user-level vs. session-level to avoid interference and ensure independence.
  • Sample size calculation: power analysis to detect a minimum detectable effect (MDE) with adequate power (e.g., 80%).
  • Primary metric: redemption rate or incremental revenue; guardrail metrics: user experience, retention, and long-term value.
  • Statistical tests: use of t-test, bootstrap, or sequential testing; correction for multiple comparisons if needed.
  • Potential pitfalls: novelty effects, network effects, and seasonality; mitigation through holdout groups and longer experiment duration.
  • Practical significance: consider business impact beyond statistical significance, such as cost of implementation and ROI.

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

Q2

How would you define the primary success metric for this experiment, and what guardrail metrics would you put in place?

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

Went with incremental profit per eligible user, which is redeem uplift times expected GMV minus the promo cost.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the specific product change being tested, then propose a primary metric that directly measures that goal and is sensitive to the change. Follow with a set of guardrail metrics that capture potential negative side effects on user experience, business health, and long-term value.

Pro tip: Emphasize that the primary metric should be a leading indicator of long-term value, not just a short-term proxy, and that guardrails should include both user-centric and business-centric metrics to catch unintended consequences.

1. Clarify experiment context and goal

Ask questions to understand the specific product change, target audience, and the business objective (e.g., increase rider engagement, improve driver utilization). This ensures the metric aligns with the intended outcome.

2. Define the primary success metric

Choose a single metric that directly measures the experiment's goal, is sensitive to the change, and is a leading indicator of long-term value. For example, if testing a new rider incentive, the primary metric could be completed trips per rider.

3. Identify potential negative side effects

Brainstorm ways the change could harm other aspects of the user experience or business, such as increased cancellations, lower driver earnings, or reduced safety.

4. Select guardrail metrics

Choose metrics that monitor these potential harms, covering user experience (e.g., cancellation rate, support tickets), business health (e.g., revenue per trip, driver retention), and long-term value (e.g., rider retention).

5. Set thresholds and monitoring plan

Define acceptable thresholds for guardrail metrics (e.g., no more than 2% degradation) and explain how you would monitor them during the experiment to make data-driven decisions.

Key Points to Mention

  • Primary metric should be a direct measure of the experiment's goal and sensitive to the change.
  • Guardrail metrics should cover user experience, business health, and long-term value.
  • Use leading indicators to predict long-term impact.
  • Consider both statistical significance and practical significance.
  • Set pre-defined thresholds for guardrails to avoid p-hacking.
  • Monitor guardrails continuously and be ready to stop the experiment if they are breached.

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

Q3

Walk through the sample size calculation for this experiment given a 3% baseline redemption rate, a 0.5 percentage point minimum detectable effect, alpha of 0.05 two-sided, and 80% power.

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

I know the formula and wrote it out fine.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the metric and test design (two-proportion z-test, equal allocation). Then compute the required sample size per variant using the standard formula for proportions, plugging in baseline p1=0.03, p2=0.035, alpha=0.05 two-sided, and power=0.80. Finally, interpret the result in practical terms (total sample, duration, and sensitivity to assumptions).

Pro tip: Mention that the sample size is driven by the smaller of the two proportions (baseline) and that you'd validate with a quick simulation or power analysis tool to account for any design nuances (e.g., unequal allocation, clustering). Also note that the MDE is absolute (0.5pp), not relative.

1. Clarify parameters and test design

Confirm the metric is a proportion (redemption rate), the test is two-sided, and allocation is equal. State the null and alternative hypotheses.

2. Choose the appropriate formula

Use the standard sample size formula for comparing two proportions: n per group = (Z_{α/2} + Z_β)^2 * (p1(1-p1) + p2(1-p2)) / (p2 - p1)^2, where p1=0.03, p2=0.035.

3. Plug in values and compute

Calculate Z_{α/2}=1.96, Z_β=0.84, p1=0.03, p2=0.035. Compute the numerator and denominator to get n per group.

4. Interpret and adjust

Round up to the nearest integer, multiply by number of variants for total sample. Discuss practical considerations like traffic, duration, and potential adjustments for unequal allocation or multiple comparisons.

Key Points to Mention

  • Two-proportion z-test for equal allocation
  • Baseline rate p1=0.03, treatment rate p2=0.035 (absolute MDE of 0.5pp)
  • Alpha=0.05 two-sided → Z_{α/2}=1.96; Power=80% → Z_β=0.84
  • Formula: n per group = (Z_{α/2}+Z_β)^2 * (p1(1-p1)+p2(1-p2)) / (p2-p1)^2
  • Result: approximately 22,000 per group (or ~44,000 total) – verify with calculation
  • Considerations: rounding up, unequal allocation, multiple testing, and practical constraints like traffic and duration

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

Q4

What analysis approach would you use, and how would you handle variance reduction and repeated looks at the data during the experiment?

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

Talked through SRM checks first, then CUPED using pre-period outcomes as the covariate.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by outlining a rigorous experimental design that includes randomization, sample size calculation, and choice of analysis method (e.g., regression adjustment, CUPED). Then explain how you would reduce variance using techniques like stratification, covariate adjustment, or CUPED, and how you would handle repeated looks with sequential testing or alpha-spending to control false positives.

Pro tip: Emphasize that variance reduction and sequential testing are not just statistical fixes but also require cross-functional alignment with product and engineering teams to ensure proper implementation and interpretation.

1. Define the experiment and metrics

Clearly state the null and alternative hypotheses, primary and guardrail metrics, and the unit of randomization. This sets the foundation for choosing the right analysis approach.

2. Choose the analysis method

Select a method that balances power and simplicity, such as regression adjustment with pre-experiment covariates or CUPED, to increase sensitivity without introducing bias.

3. Implement variance reduction techniques

Apply techniques like stratification, covariate adjustment, or CUPED to reduce variance and increase the chance of detecting a true effect. Explain how these methods work and their assumptions.

4. Address repeated looks at the data

Use sequential testing methods (e.g., group sequential boundaries, alpha-spending functions) or always-valid inference to control the overall false positive rate when monitoring results over time.

5. Validate and interpret results

Check for SRM, ensure assumptions are met, and interpret the effect size with confidence intervals. Discuss practical significance and potential trade-offs.

Key Points to Mention

  • Randomization and sample size calculation
  • CUPED (Controlled-experiment Using Pre-Experiment Data) or regression adjustment
  • Stratification or blocking to reduce variance
  • Sequential testing (e.g., O'Brien-Fleming boundaries, alpha-spending)
  • False discovery rate control and multiple testing corrections
  • Practical significance vs. statistical significance

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

Q5

How would you handle interference between users or channel saturation in this experiment, and what alternative designs would you consider?

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

This is where I felt most underprepared.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by acknowledging that interference and saturation are common in marketplace experiments like Uber's, then explain how you would diagnose and mitigate them using techniques like cluster randomization or switchback designs. Emphasize the trade-offs between bias reduction and practical constraints, and propose alternative designs with clear reasoning.

Pro tip: Show that you understand Uber's two-sided marketplace dynamics: interference often arises from supply-demand interactions, so consider designs that account for both riders and drivers, such as bipartite randomization or time-based switchbacks.

1. Diagnose interference and saturation

Identify symptoms like spillover effects, network effects, or resource competition. Use metrics such as treatment effect heterogeneity across geographies or time to detect interference.

2. Choose mitigation strategies

Apply techniques like cluster randomization (e.g., by city or driver), switchback designs, or budget splitting to reduce interference. Consider using holdout groups or synthetic controls.

3. Evaluate trade-offs

Discuss how each design affects statistical power, bias, and operational feasibility. For example, cluster randomization reduces interference but increases variance.

4. Propose alternative designs

Suggest alternatives like switchback, cluster randomized, or bipartite experiments. Explain when each is appropriate and how they address saturation.

5. Validate and iterate

Recommend running A/A tests or simulations to validate design assumptions. Be prepared to adapt based on results and business constraints.

Key Points to Mention

  • Interference types: spillover, network effects, resource competition (e.g., driver supply).
  • Saturation: when treatment exhausts shared resources, leading to biased estimates.
  • Cluster randomization: randomize at group level (e.g., city) to contain interference.
  • Switchback designs: alternate treatment/control over time to mitigate interference in marketplaces.
  • Trade-offs: bias vs. variance, power, and operational complexity.
  • Alternative designs: bipartite randomization, synthetic control, or holdout groups.

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

Q6

How would you ensure budget parity between experiment arms over time, and how do you handle throttling, suppression lists, and spillover?

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

Practical ops question that I wasn't expecting to go this deep on.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the goal: maintaining valid comparisons by ensuring balanced assignment and minimizing interference. Then walk through mechanisms for budget parity (e.g., sequential testing, balanced allocation), throttling (e.g., rate limiting, traffic shaping), suppression lists (e.g., exclusion criteria, holdouts), and spillover (e.g., cluster randomization, geo-based experiments). Emphasize trade-offs and practical implementation at scale.

Pro tip: Highlight that perfect parity is often impossible; instead, focus on detecting and correcting imbalances via monitoring and using techniques like CUPED or stratified randomization to increase sensitivity. Also, mention that spillover can be mitigated by using switchback or cluster designs, which are common at Uber.

1. Define the problem and constraints

Clarify what budget parity means in context (e.g., equal sample sizes, equal exposure, or equal spend) and identify potential sources of imbalance and interference.

2. Ensure balanced allocation over time

Use randomization units (user, session, geo) and techniques like stratified randomization or sequential testing to maintain balance as data accumulates.

3. Implement throttling and suppression

Apply throttling to control traffic volume and avoid overload; use suppression lists to exclude users (e.g., those already exposed, employees) and maintain clean comparisons.

4. Mitigate spillover effects

Choose appropriate randomization units (e.g., clusters, geos) and consider switchback designs to reduce contamination between arms.

5. Monitor and adjust

Continuously monitor key metrics for imbalance and interference; use statistical methods to detect and correct issues, and be prepared to restart or adjust the experiment.

Key Points to Mention

  • Randomization unit selection (user vs. session vs. geo) and its impact on spillover
  • Sequential testing and alpha spending to allow continuous monitoring without inflating false positives
  • Throttling mechanisms: rate limiting, traffic shaping, and their effect on experiment power
  • Suppression lists: exclusion criteria, holdout groups, and ensuring they are applied consistently
  • Spillover mitigation: cluster randomization, switchback experiments, and network effects
  • Statistical techniques to adjust for imbalances: CUPED, stratification, regression adjustment

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