← DoorDash Interview Insights

DoorDash·Data Scientist·Technical Phone Screen·Senior

SeniorPrefer not to say
Jun 2026Remote

Summary

DoorDash data scientist interview that went deep on experimentation design and causal inference. The whole session was basically one long scenario about a recommendation algorithm, which sounds focused but actually covers a lot of ground fast.

Questions Asked (3)

Q1

Design an A/B test to evaluate a new recommendation algorithm's impact on revenue. Walk through your hypotheses, how you'd pick the unit of randomization, sample size calculation, and what metrics you'd track.

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

This part felt manageable until I got to unit of randomization.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the business goal and articulating clear hypotheses about how the new recommendation algorithm will affect revenue. Then walk through the experimental design—randomization unit, sample size, and metrics—while emphasizing practical considerations like network effects and guardrail metrics. Conclude with how you'd analyze results and make a recommendation.

Pro tip: At DoorDash, randomization by user can be tricky due to shared accounts and device switching; consider randomizing by session or device ID, and always check for interference between treatment and control groups. Also, define revenue metrics that align with DoorDash's marketplace dynamics, such as orders per user and average order value.

1. Define Hypotheses

State a clear null and alternative hypothesis: e.g., the new algorithm increases revenue per user compared to the current algorithm. Break down into primary and secondary hypotheses (e.g., impact on order frequency, basket size).

2. Choose Randomization Unit

Decide whether to randomize by user, session, or store. Consider trade-offs: user-level avoids contamination but may dilute effect if algorithm is session-based; session-level increases power but risks spillover. At DoorDash, account for shared devices and multiple sessions per user.

3. Calculate Sample Size

Determine required sample size using power analysis: specify significance level (α=0.05), power (1-β=0.8), minimum detectable effect (MDE) based on business relevance, and baseline variance. Use historical data to estimate variance and consider clustering if randomizing by store.

4. Select Metrics

Choose primary metric: revenue per user (or per session). Include secondary metrics: conversion rate, average order value, order frequency. Add guardrail metrics: customer satisfaction, delivery time, cancellation rate, and long-term retention.

5. Analyze and Interpret

Plan for statistical analysis: use t-tests or regression with covariates, check for novelty effects, and segment by user type. Consider sequential testing or Bayesian methods if peeking. Finally, assess practical significance and recommend rollout if results are positive and guardrails are intact.

Key Points to Mention

  • Randomization unit trade-offs: user vs. session vs. store, and how to handle interference
  • Sample size calculation inputs: baseline conversion rate, MDE, power, significance level, and variance
  • Primary metric: revenue per user; secondary metrics: orders, AOV, conversion; guardrails: cancellations, delivery time, customer satisfaction
  • Potential pitfalls: network effects, novelty effect, multiple testing, and seasonality
  • Analysis techniques: intention-to-treat, covariate adjustment, and heterogeneous treatment effects
  • Business context: DoorDash's three-sided marketplace and how algorithm changes affect restaurants and dashers

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

Q2

Your A/B test comes back with p = 0.08 for revenue lift. How do you interpret that and would you ship the feature?

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

I said no ship, p > 0.05, end of story.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by explaining that p=0.08 means the result is not statistically significant at the conventional 0.05 threshold, but it doesn't mean the feature has no effect. Then discuss the business context: the potential revenue lift, the cost of shipping, and the risk of false positives. Finally, recommend a decision based on a cost-benefit analysis and possibly suggest running the test longer or segmenting the data.

Pro tip: Mention that p-values are not the only factor; consider confidence intervals, practical significance, and the expected value of shipping. Also, at a company like DoorDash, emphasize the importance of aligning with business goals and avoiding premature decisions based on a single metric.

1. Interpret the p-value

Explain that p=0.08 means there is an 8% chance of observing such an extreme result if the null hypothesis (no effect) were true. It does not prove the null; it's suggestive but not conclusive at the 0.05 level.

2. Assess practical significance

Look at the effect size and confidence interval. Even if not statistically significant, the lift might be practically meaningful. Consider the confidence interval: does it include zero? What is the range of plausible effects?

3. Consider business context

Evaluate the potential revenue impact, implementation costs, and risks. For DoorDash, consider factors like market competitiveness, user experience, and long-term effects. Also, check if the test was adequately powered.

4. Make a decision

Decide whether to ship, iterate, or run a follow-up test. If the expected value is positive and the risk is low, shipping might be justified. Otherwise, gather more data or refine the feature.

Key Points to Mention

  • p-value interpretation and limitations
  • Statistical significance vs. practical significance
  • Confidence intervals and effect size
  • Power analysis and sample size
  • Business impact and cost-benefit analysis
  • Potential for false positives and multiple testing

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

Q3

If you couldn't run a randomized experiment, how would you estimate the causal impact of the algorithm? Compare difference-in-differences, propensity score matching, and instrumental variables.

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

Honestly the part I was least prepared for.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the core challenge of estimating causal impact without randomization, then compare the three methods on key dimensions: assumptions, data requirements, and applicability to DoorDash's marketplace. Use a concrete example (e.g., a new dispatch algorithm) to illustrate when each method is appropriate and how you would validate results.

Pro tip: Emphasize that no observational method is perfect; the best approach often combines multiple methods and triangulates results. Also, mention that in a marketplace like DoorDash, interference and dynamic pricing can violate key assumptions, so you must carefully check for SUTVA violations and consider cluster-level or switchback designs.

1. Clarify the causal question and constraints

Define the treatment (e.g., new algorithm), outcome (e.g., delivery time), and population. Explain why randomization is infeasible (e.g., ethical, logistical, or marketplace interference).

2. Explain each method's logic and assumptions

For DiD: parallel trends; for PSM: conditional independence (no unmeasured confounders); for IV: relevance and exclusion restriction. Briefly describe how each estimates the causal effect.

3. Compare methods on data needs and validity

Discuss data requirements (panel data for DiD, rich covariates for PSM, a valid instrument for IV) and threats to validity (e.g., time-varying confounders, poor overlap, weak instruments).

4. Apply to DoorDash context

Give a concrete example: e.g., estimating the impact of a new dispatch algorithm on delivery time. Discuss which method might work best given available data and marketplace dynamics.

5. Recommend a practical approach

Suggest combining methods (e.g., DiD with PSM to adjust for covariates) and validating with placebo tests, sensitivity analyses, or if possible, a switchback experiment.

Key Points to Mention

  • Difference-in-differences: requires parallel trends assumption; can control for time-invariant confounders; useful when treatment is rolled out to some regions at different times.
  • Propensity score matching: balances observed covariates; relies on conditional independence; can be combined with DiD to form a doubly robust estimator.
  • Instrumental variables: requires a valid instrument that affects treatment but not outcome except through treatment; often hard to find in practice; can estimate local average treatment effect (LATE).
  • Marketplace interference: in platforms like DoorDash, treatment of one unit can affect others (SUTVA violation), making these methods challenging; consider cluster-level or switchback designs.
  • Sensitivity analysis: assess robustness to unmeasured confounding (e.g., Rosenbaum bounds for PSM, placebo tests for DiD, weak instrument diagnostics for IV).
  • Triangulation: use multiple methods and see if results converge; if not, investigate why and refine assumptions.

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