Amazon Business Analyst Interview Questions (2026)

1. 1Online Assessment (SDE OA)60 minMedium

   Two timed coding problems plus a workplace-simulation and logic section; the main gate for freshers and India volume hiring.

2. 2Phone screen45 minMedium

   One coding problem plus 1-2 Leadership Principle STAR questions with an SDE.

3. 3Coding loop round60 minMedium

   DSA problem to working code, followed by assigned-LP behavioral questions in STAR format.

4. 4System design loop round60 minHard

   Design an Amazon-scale service with capacity math, plus LPs; low-level/OOD design substitutes for junior candidates.

5. 5Hiring Manager round45 minMedium

   Team fit, project deep dives, and Deliver Results/Bias for Action stories with the manager you would report to.

6. 6Bar Raiser60 minHard

   An interviewer from outside the team stress-tests LP stories and overall bar with the hardest cross-examination of the loop; holds veto.

1. Q1

   Design an A/B test for a new Retail Marketplace ranking or recommendation change. Define hypothesis, primary metric, guardrails, randomization unit, and launch decision rule

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Context: Amazon wants to grow retail conversion while keeping delivery promises and inventory healthy.

   How to answer: A strong answer will define a clear, testable hypothesis for the ranking/recommendation change, such as 'The new algorithm will increase customer engagement.' The primary metric should directly reflect the hypothesis, like 'Click-Through Rate (CTR) on recommended items' or 'Conversion Rate for items from the new ranking block.' Guardrail metrics are crucial, including 'Average Order Value (AOV),' 'Page Load Time,' and 'Customer Service Contact Rate,' to ensure no negative impact on other key business areas. The randomization unit should be 'individual user' or 'session' to minimize contamination. Finally, the launch decision rule should specify statistical significance (e.g., p < 0.05) and practical significance (e.g., minimum 1% uplift in primary metric, with no significant negative impact on guardrails) over a defined test duration.

2. Q2

   For Prime, should randomization happen at customer, session, device, seller, or region level? Explain the tradeoffs

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Consider cross-device behavior, interference, marketplace effects, and operational feasibility.

   How to answer: Randomization for Prime A/B testing should typically happen at the customer level to ensure independent observations and avoid contamination, as Prime benefits are tied to a customer's account. However, session or device level might be considered for very specific, short-lived UI/UX tests where the change is not persistent across logins or devices. Seller or region level randomization would be appropriate for tests impacting supply-side changes, pricing strategies, or logistical experiments, but not for core Prime membership features. The key is to choose the level that minimizes contamination and maximizes statistical power for the specific hypothesis being tested.

3. Q3

   Choose primary and guardrail metrics for a Buy Box experiment aimed at improving purchase conversion rate. What metrics would prevent a harmful launch?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Include user experience, partner health, revenue, reliability, and long-term retention considerations.

   How to answer: For a Buy Box experiment focused on improving purchase conversion, the primary metric would be 'Purchase Conversion Rate' (units purchased / sessions with product view). Guardrail metrics are crucial to prevent negative impacts. Key guardrails include 'Revenue per Session' to ensure increased conversions don't come from lower-value purchases, 'Average Selling Price' to monitor product mix, and 'Customer Satisfaction' (e.g., return rate, customer service contacts) to ensure the experience isn't frustrating. Additionally, 'Page Load Time' and 'Error Rate' are important technical guardrails.

4. Q4

   During a Sponsored Products experiment, the treatment/control split is 52/48 instead of 50/50. How would you diagnose sample ratio mismatch?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Assume assignment logs, exposure logs, and eligibility filters may disagree.

   How to answer: A strong candidate would first acknowledge the 52/48 split as a potential Sample Ratio Mismatch (SRM) and state its importance. They would then outline a diagnostic process starting with checking the implementation of the random assignment mechanism (e.g., hashing function, bucketing logic, user ID consistency). Next, they would investigate potential data pipeline issues such as data logging errors, ETL discrepancies, or delayed data ingestion that could skew counts. Finally, they would consider pre-existing user characteristics or platform interactions that might lead to differential exposure or drop-off rates between groups, even with correct assignment.

5. Q5

   The Alexa Shopping experiment is trending positive after two days. A PM wants to stop early and launch. How do you handle peeking and sequential testing?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Discuss pre-specified stopping rules, alpha spending, business urgency, and risk.

   How to answer: A strong candidate would first explain the dangers of peeking, specifically inflated Type I error rates (false positives). They would then detail methods to mitigate this, such as pre-determining sample size and experiment duration based on Minimum Detectable Effect (MDE), statistical power, and significance level. If early stopping is truly desired, they would discuss sequential testing methodologies like Always Valid P-values (AVP) or using O'Brien-Fleming boundaries, which adjust the significance threshold over time to maintain the overall Type I error rate. Finally, they would emphasize the importance of leadership alignment and a clear decision-making framework for A/B testing at Amazon.

6. Q6

   A new Fulfillment by Amazon feature shows a large week-1 lift in purchase conversion rate, but the effect fades by week 4. What could explain this and how would you design the test duration?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Discuss novelty, learning effects, seasonality, and durable impact.

   How to answer: The fading lift could be due to novelty effect, where initial user engagement is high but normalizes as the feature becomes standard. Another explanation is that the feature primarily benefits new users or specific use cases that are exhausted over time, or that a confounding variable (e.g., promotional activity) coincided with the initial launch. To design test duration, I would consider the typical customer lifecycle, seasonality, and sufficient time for user behavior to stabilize and for any long-term negative effects (e.g., increased returns) to manifest. A minimum of 4-6 weeks is often a good starting point, with continuous monitoring for stabilization.

7. Q7

   In a marketplace-like Retail Marketplace feature, treatment users may affect control users. How would network effects or interference bias the experiment?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Examples include seller supply, content inventory, delivery capacity, or pricing pressure.

   How to answer: Network effects in a marketplace A/B test occur when the treatment group's actions influence the control group's experience, leading to interference. This typically biases the experiment by understating the true treatment effect if the interference is negative (e.g., treatment users consume limited resources, harming control users). Conversely, positive interference (e.g., treatment users create more supply, benefiting control users) could overstate the treatment effect in a simple A/B test. The bias arises because the control group no longer represents a true baseline without the treatment, making direct comparison inaccurate.

8. Q8

   purchase conversion rate is a low-frequency event for Prime. How would you set up an experiment with enough power without waiting too long?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Discuss proxy metrics, variance reduction, larger samples, longer windows, and risk of metric gaming.

   How to answer: To address low-frequency events like Prime purchase conversion, consider using a proxy metric that is a leading indicator and has higher frequency, such as 'add to cart' or 'view product details' for Prime-eligible items. Alternatively, expand the experiment's scope to include more users or a longer duration, but carefully weigh the trade-offs in terms of risk and time. Employ sequential testing or Bayesian methods to potentially stop the experiment early if a clear winner emerges, or use a composite metric combining several related, higher-frequency actions. Finally, ensure the minimum detectable effect (MDE) is realistically set, as a smaller MDE requires more data.

9. Q9

   Design a geo or region-level experiment for Buy Box. When is this better than user-level randomization, and what are the analytical downsides?

   MediumStatistics & Experimentation RoundA/B TestingAmazon-specific

   Context: Use matched markets, pre-period balancing, spillover checks, and fewer experimental units.

   How to answer: A geo-level experiment for Buy Box involves randomizing entire geographic regions (e.g., states, cities, or postal codes) into treatment and control groups, rather than individual users. This approach is superior to user-level randomization when there's a high risk of network effects or contamination, such as when changes to the Buy Box algorithm might influence seller behavior across a region, which in turn affects other buyers. Analytical downsides include reduced statistical power due to fewer, larger units of randomization, increased sensitivity to selection bias if regions aren't perfectly balanced, and longer experiment durations to achieve significance.

10. Q10

    The Sponsored Products experiment lifts purchase conversion rate overall, but only for new users and only in one category. How would you evaluate heterogeneous treatment effects?

    HardStatistics & Experimentation RoundA/B TestingAmazon-specific

    Context: Balance pre-planned segments with exploratory slicing and multiple testing risk.

    How to answer: A strong candidate would first acknowledge the initial A/B test results and the need for deeper segmentation. They would propose using statistical methods like interaction terms in regression models or CUPED to isolate and quantify the treatment effect for specific user segments (new vs. existing) and product categories. Further, they would discuss the importance of power analysis for these segmented analyses, ensuring sufficient sample size within each subgroup to detect meaningful effects. Finally, they would recommend follow-up qualitative research or targeted experiments to understand the 'why' behind the observed heterogeneity.

11. Q11

    Treatment improves purchase conversion rate but worsens on-time delivery rate for Alexa Shopping. Walk through a launch recommendation

    HardStatistics & Experimentation RoundA/B TestingAmazon-specific

    Context: Make a decision under conflicting metrics and quantify tradeoffs for stakeholders.

    How to answer: A strong recommendation balances the trade-offs between improved purchase conversion and worsened on-time delivery. The candidate should first quantify the impact of both metrics (e.g., revenue uplift vs. potential customer churn/support costs) and identify the target customer segment. They must then propose a phased rollout strategy, starting with a small, controlled group to gather more data on long-term customer behavior and sentiment, especially regarding delivery reliability. Finally, the recommendation should include a plan for mitigating the negative impact on delivery, such as optimizing logistics or offering delivery incentives, before a full-scale launch.

12. Q12

    How would you design ramp-up, holdback, and post-launch monitoring for a successful Fulfillment by Amazon A/B test?

    HardStatistics & Experimentation RoundA/B TestingAmazon-specific

    Context: Include ramp stages, persistent holdback, alert thresholds, rollback criteria, and owner accountability.

    How to answer: For ramp-up, I'd start with a small percentage (e.g., 1-5%) of traffic, closely monitoring key operational metrics like latency, error rates, and fulfillment success rates, not just A/B test metrics. I'd gradually increase exposure (e.g., 10%, 25%, 50%, 100%) only after confirming stability and performance at each stage. Holdback involves reserving a small, consistent percentage of traffic (e.g., 1-2%) on the original experience for an extended period post-launch to serve as a long-term control, allowing detection of latent or seasonal effects. Post-launch monitoring requires ongoing vigilance on both business metrics (e.g., delivery speed, defect rates, customer contacts) and technical health, with automated alerts and dashboards, comparing the new experience against the holdback and pre-launch baselines.

13. Q13

    Midway through the Retail Marketplace test, tracking for Buy Box changed. How would you decide whether the experiment results are still usable?

    HardStatistics & Experimentation RoundA/B TestingAmazon-specific

    Context: Compare instrumentation versions, affected traffic share, raw logs, and sensitivity analyses.

    How to answer: First, assess the nature and timing of the Buy Box tracking change. Determine if the change impacts the definition or measurement of the Buy Box metric itself, and whether it occurred before or after a significant portion of the experiment data was collected. If the change is a minor backend fix or occurred very early, results might be salvageable with careful analysis. However, if the change alters the metric's meaning or introduces bias mid-experiment, the experiment is likely compromised and should be restarted or re-evaluated with extreme caution, potentially focusing on other unaffected metrics.

14. Q14

    Two overlapping experiments on Prime both affect gross merchandise sales. How would you detect and manage interaction effects?

    HardStatistics & Experimentation RoundA/B TestingAmazon-specific

    Context: Discuss experiment registry, factorial design, exclusion rules, and interaction terms.

    How to answer: A strong candidate would first emphasize the importance of pre-experiment design to minimize overlap, perhaps through segmentation or sequential rollout. If overlap is unavoidable, they would propose a factorial experimental design (e.g., 2x2 matrix) to explicitly test for interaction effects by observing the combined impact versus the sum of individual impacts. Analysis would involve statistical tests (e.g., ANOVA) to determine if the interaction term is significant. If significant, they would recommend either redesigning the experiments, sequential rollout, or accepting the interaction and modeling its impact for future decision-making.

15. Q15

    Amazon's Retail Marketplace revenue suddenly drops 10% week over week. Structure a business case to diagnose the issue and identify the most likely drivers

    MediumProduct Analytics & Business CaseBusiness CasesAmazon-specific

    Context: Consider traffic, conversion, pricing, mix, supply/inventory, outages, marketing, and seasonality.

    How to answer: A strong business case would begin by clarifying the scope (e.g., specific marketplace, product categories, geographies) and establishing a baseline. The diagnosis should follow a structured approach, starting with internal factors (e.g., system outages, pricing errors, inventory issues, policy changes) and then moving to external factors (e.g., competitor actions, economic shifts, seasonality, major news events). Prioritize potential drivers based on impact and likelihood, leveraging data (e.g., traffic, conversion, ASP, seller activity) to validate hypotheses. Finally, recommend specific actions to mitigate the revenue drop and monitor the effectiveness of those interventions.

Practice these with instant AI feedback in a live mock interview → Start a Amazon Business Analyst mock

Prepare 8-12 STAR stories mapped to Leadership Principles; expect a Bar Raiser; quantify impact

Indicative Business Analyst pay in India: ~₹7–26 LPA (role-level range, not a Amazon-specific figure).

Other roles at Amazon

Business Analyst interviews at other companies