← Lyft Interview Insights

Lyft·Data Scientist·Onsite - Multi Round·Senior

Senior
Jun 2026

Summary

Lyft data science onsite with a very heavy experimentation focus. The whole thing was essentially one long case about redesigning supplier prioritization in a two-sided marketplace, broken into five parts that kept building on each other. Tough but fair.

Questions Asked (6)

Q1

You're proposing a new supplier ranking policy to boost order completion in a two-sided marketplace with known interference between users and suppliers. How would you design the experiment, and which randomization unit would you choose given network effects and spillover risks?

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

This is where I spent the most time and also where I fumbled first.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the interference problem: in a two-sided marketplace, supplier ranking affects both riders and drivers, so user-level randomization will be biased by spillovers. Propose a cluster-based randomization (e.g., by city or driver cohort) with a switchback or staggered rollout design, and define clear metrics for order completion and marketplace balance.

Pro tip: Emphasize that you would run a power analysis accounting for intra-cluster correlation and consider a holdout group to measure long-term effects, showing you understand the trade-offs between bias and variance in networked experiments.

1. Identify interference and spillover mechanisms

Map how the new ranking policy could affect both riders and drivers, and how interactions between them (e.g., driver repositioning, rider wait times) create spillovers across units.

2. Choose randomization unit and design

Select a unit that minimizes contamination, such as geographic clusters (cities or neighborhoods) or driver cohorts, and consider switchback or staggered rollout to balance bias and power.

3. Define metrics and guardrails

Specify primary metrics (e.g., order completion rate, ETA, match rate) and guardrails (e.g., driver utilization, rider cancellation) to capture both sides of the marketplace.

4. Plan analysis and power

Account for clustering in power calculations, use appropriate statistical methods (e.g., cluster-robust standard errors, CUPED), and pre-register the analysis plan.

5. Validate and iterate

Run a pilot to check for spillovers, monitor for novelty effects, and be prepared to adjust the design if interference is stronger than expected.

Key Points to Mention

  • Network effects and spillover risks in two-sided marketplaces
  • Randomization units: user vs. driver vs. geographic cluster vs. switchback
  • Trade-offs between bias and variance in cluster randomization
  • Metrics for both sides: order completion, ETA, driver acceptance, rider wait time
  • Power analysis with intra-cluster correlation and minimum detectable effect
  • Potential use of holdout groups or staggered rollout to measure long-term effects

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

Q2

What primary metric and guardrails would you define for this experiment, and at what level would you aggregate them?

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

Felt more comfortable here.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the product surface (e.g., rider app, driver app, marketplace). Then propose a primary metric that directly measures the intended outcome, along with guardrail metrics that protect user experience and business health. Finally, specify the aggregation level (e.g., per rider, per driver, per city) and justify it based on the unit of randomization and the metric's sensitivity.

Pro tip: Always tie the choice of primary metric and guardrails to the experiment's hypothesis and the company's north star. For Lyft, consider marketplace dynamics: a change that boosts rider conversions might hurt driver utilization, so include both sides as guardrails.

1. Clarify experiment goal and hypothesis

Ask or infer what the experiment is trying to change (e.g., increase rider bookings, reduce driver wait times). This determines the primary metric.

2. Select primary metric

Choose a single metric that directly measures the desired outcome and is sensitive to the change. For Lyft, examples: rides per rider, driver acceptance rate, or ETAs.

3. Define guardrail metrics

Identify metrics that should not degrade, such as rider cancellations, driver earnings, or system latency. Include both user experience and business health metrics.

4. Determine aggregation level

Decide whether to aggregate at the user, driver, city, or trip level. This should align with the randomization unit and the metric's natural unit of analysis.

5. Justify and validate choices

Explain why these metrics and aggregation levels are appropriate, and mention any trade-offs or potential pitfalls (e.g., network effects, seasonality).

Key Points to Mention

  • Primary metric should be directly tied to the experiment's hypothesis and be a leading indicator of long-term value.
  • Guardrails should cover both user experience (e.g., cancellations, ratings) and business metrics (e.g., revenue, driver utilization).
  • Aggregation level must match the randomization unit to avoid bias; for Lyft, often user-level or city-level.
  • Consider marketplace two-sidedness: changes affecting riders may impact drivers, so include metrics for both sides.
  • Mention statistical power and minimum detectable effect to ensure the experiment can detect meaningful changes.
  • Discuss potential novelty effects and how to monitor them via guardrails over time.

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 test. Baseline completion rate is 60%, minimum detectable effect is +2 percentage points absolute, two-sided test at alpha 0.05 and 80% power. Then adjust for a clustering design effect of 1.2.

A/B Testing & Experimentation
Author's notes

I actually like power calculation questions because they're concrete.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the test design (e.g., unit of randomization, metric type) and then walk through the standard sample size formula for a two-proportion z-test, plugging in the given values. After obtaining the base sample size, multiply by the design effect to account for clustering, and finally discuss any additional considerations like unequal allocation or multiple comparisons.

Pro tip: Always state the formula and assumptions explicitly, and mention that the design effect inflates the variance, so you multiply the sample size by the design effect (not the variance). Also, note that the calculation assumes equal allocation and no other adjustments; in practice, you'd round up and account for other factors like non-compliance.

1. Clarify parameters and assumptions

Confirm the baseline rate (p1 = 0.60), minimum detectable effect (absolute +2 percentage points, so p2 = 0.62), two-sided alpha = 0.05, power = 0.80, and design effect = 1.2. State that you assume equal allocation and independent observations (before clustering adjustment).

2. Compute base sample size per group

Use the standard formula for comparing two proportions: n per group = ( (z_{1-α/2} + z_{1-β})^2 * (p1(1-p1) + p2(1-p2)) ) / (p2 - p1)^2. Plug in z_{0.975} ≈ 1.96, z_{0.80} ≈ 0.84, and the proportions to calculate n.

3. Adjust for clustering design effect

Multiply the base sample size per group by the design effect (1.2) to account for the increased variance due to clustering. This yields the required sample size per group under the clustered design.

4. Finalize and discuss practical considerations

Round up to the nearest integer, and mention that you might further adjust for unequal allocation, expected non-compliance, or multiple testing. Also, clarify that the total sample size is twice the per-group size if equal allocation.

Key Points to Mention

  • Use the two-proportion z-test formula for sample size calculation.
  • Specify z-scores: 1.96 for 95% confidence (two-sided) and 0.84 for 80% power.
  • Calculate the base sample size per group before design effect.
  • Multiply by design effect (1.2) to adjust for clustering.
  • Round up to ensure sufficient power.
  • Discuss assumptions: equal allocation, independence (pre-adjustment), and no other adjustments.

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

Q4

How would you handle heterogeneous treatment effects across cities, time of day, and supplier capacity quartiles? And how do you run sequential monitoring without inflating Type I error?

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

Two questions crammed into one, which I think was intentional.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the problem as a need to balance overall treatment effect estimation with subgroup insights, using pre-registered heterogeneity analysis to avoid false positives. Then discuss sequential monitoring methods like alpha spending or group sequential designs to control Type I error while allowing early stopping. Emphasize practical implementation at Lyft, including logging, automation, and decision-making.

Pro tip: Pre-register your heterogeneity hypotheses and use a hierarchical model to borrow strength across subgroups, which reduces false positives and increases power. For sequential monitoring, use a validated alpha-spending function like O'Brien-Fleming and always adjust for multiple comparisons across subgroups.

1. Define hypotheses and subgroups

Clearly specify the primary metric and pre-register which dimensions (city, time of day, supplier capacity quartile) you will analyze for heterogeneity, along with the expected direction of effects.

2. Choose appropriate statistical methods

Use interaction terms in regression models or hierarchical Bayesian models to estimate subgroup effects while controlling for multiple comparisons. Consider causal forest for exploratory analysis.

3. Implement sequential monitoring

Apply group sequential designs or alpha-spending functions (e.g., O'Brien-Fleming, Pocock) to allow interim analyses without inflating Type I error. Use software like gsDesign or sequential package.

4. Adjust for multiple testing

Apply corrections like Bonferroni, Holm, or false discovery rate (FDR) for subgroup analyses, and ensure the sequential monitoring plan accounts for all interim looks.

5. Operationalize and communicate

Automate monitoring dashboards, define stopping rules, and communicate results with confidence intervals and effect sizes, emphasizing practical significance over statistical significance.

Key Points to Mention

  • Pre-registration of subgroups and analysis plan to avoid p-hacking
  • Use of interaction terms or hierarchical models for heterogeneous treatment effects
  • Sequential monitoring methods: alpha spending, group sequential designs, or always-valid inference
  • Multiple testing corrections (Bonferroni, Holm, FDR) for subgroup analyses
  • Practical considerations: sample size, power, and decision-making trade-offs
  • Tools and implementation: gsDesign, sequential, or Bayesian approaches

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

Q5

How would you detect and correct for marketplace rebalancing effects, like treated geos improving at the expense of control geos?

A/B Testing & ExperimentationRoot Cause Analysis
Author's notes

Genuinely did not see this one coming in the way it was framed.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by acknowledging that marketplace rebalancing is a form of interference where treatment affects control units, violating SUTVA. Then outline a diagnostic approach to detect it, such as comparing pre-period trends and checking for negative treatment effects in control, and propose correction methods like cluster-based randomization, switchback designs, or modeling the interference explicitly.

Pro tip: In marketplace experiments, always monitor the ratio of treated to control units in each geographic area and look for spillover effects; if detected, consider using a difference-in-differences approach with synthetic controls to isolate the direct effect.

1. Detect potential rebalancing

Compare pre-experiment trends between treatment and control geos, and check for unexpected changes in control metrics (e.g., a decline) that suggest treatment is drawing activity away. Use metrics like supply/demand balance, wait times, and conversion rates.

2. Quantify interference

Measure the magnitude of spillover by analyzing cross-geo correlations or using a model that accounts for interference, such as a spatial regression or a marketplace simulation. Estimate the bias in treatment effect.

3. Choose correction method

Select an appropriate design or analysis method: cluster randomization (e.g., by city), switchback experiments, or use of instrumental variables. Alternatively, model the interference directly with techniques like causal inference under interference.

4. Validate and adjust

After applying a correction, re-run diagnostics to ensure interference is mitigated. Compare results from different methods to check robustness, and if possible, run a holdout or validation experiment.

5. Communicate findings

Clearly explain the detected interference, the chosen correction, and the adjusted effect size to stakeholders, highlighting any remaining uncertainty and recommendations for future experiments.

Key Points to Mention

  • SUTVA violation and interference in marketplace experiments
  • Use of cluster randomization or switchback designs to mitigate spillover
  • Difference-in-differences and synthetic control methods for correction
  • Monitoring supply/demand metrics and cross-geo correlations
  • Importance of pre-period trend analysis and placebo tests
  • Trade-offs between bias reduction and precision when correcting for interference

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

Q6

Design a rollout ramp plan for this policy change and define a fallback decision rule if guardrail metrics breach for two consecutive days.

A/B Testing & ExperimentationProduct Strategy
Author's notes

Saved the most operational question for last.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining the guardrail metrics and their thresholds, then outline a phased rollout with increasing traffic percentages and clear success criteria. Specify a fallback decision rule that triggers if any guardrail metric breaches for two consecutive days, including immediate actions like pausing the rollout and reverting to the previous version.

Pro tip: Emphasize the importance of pre-registering the analysis plan and guardrail thresholds to avoid p-hacking and ensure statistical validity. Also, mention the need for automated monitoring and alerting to detect breaches in real-time.

1. Define guardrail metrics and thresholds

Identify key guardrail metrics (e.g., crash rate, cancellation rate, driver acceptance rate) and set acceptable thresholds based on historical data and business impact.

2. Design phased rollout plan

Plan a gradual rollout (e.g., 1%, 5%, 10%, 25%, 50%, 100%) with predefined durations and sample sizes to detect issues early while minimizing risk.

3. Establish monitoring and alerting

Set up real-time dashboards and automated alerts to track guardrail metrics daily and notify stakeholders of any breaches.

4. Define fallback decision rule

Specify that if any guardrail metric breaches its threshold for two consecutive days, the rollout will be paused, and the change will be reverted to the previous version.

5. Document and communicate plan

Write a detailed rollout and fallback plan, share with cross-functional teams, and ensure alignment on roles and responsibilities during execution.

Key Points to Mention

  • Selection of guardrail metrics that are sensitive to the policy change and aligned with business goals.
  • Statistical power and sample size considerations for each rollout phase.
  • Use of sequential testing or Bayesian methods to allow for early stopping.
  • Automated monitoring systems and alerting thresholds to detect breaches promptly.
  • Clear escalation path and decision-making authority for fallback actions.
  • Post-mortem analysis and iterative improvement after a fallback event.

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