← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Senior

Senior
Apr 2026

Summary

A product-heavy DS interview at Meta, this one was essentially a full case study on whether you'd ship a feature to Instagram based on Facebook data. Dense, multi-part, and the kind of question where you can feel yourself running out of runway halfway through.

Questions Asked (5)

Q1

A 'More like this' button that surfaces similar products exists on Facebook but hasn't launched on Instagram. You have access to interactions and product schemas but no button telemetry on Instagram and can't use revenue as your goal. How would you build pre-launch evidence using existing data to convince a PM this feature is worth building?

Product Analytics & MetricsA/B Testing & ExperimentationProduct Sense & Ideation
Author's notes

I tried to anchor on exploration rate as my leading indicator, basically sessions where a user touched at least one similar-product interaction divided by total sessions.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Frame the problem as a causal inference question: use Facebook's 'More like this' as a natural experiment to estimate its impact on engagement, then validate that Instagram's product and interaction data can support a similar effect. Propose a pre-launch measurement plan that uses proxy metrics and quasi-experimental methods to build a compelling case for the PM.

Pro tip: Emphasize that you're not just looking for correlation but trying to establish causality by leveraging Facebook's launch as a natural experiment, and always tie your metrics back to a clear product hypothesis about user intent.

1. Define the hypothesis and success metrics

Articulate the product hypothesis: 'More like this' increases discovery and engagement by reducing search friction. Define proxy success metrics (e.g., click-through rate on similar items, session depth, time to next interaction) that are measurable on Instagram without revenue.

2. Leverage Facebook as a natural experiment

Use Facebook's launch as a quasi-experiment: compare user engagement before and after the button's introduction, and use difference-in-differences or synthetic control to isolate the button's effect. Ensure comparability between Facebook and Instagram user behaviors.

3. Validate with Instagram's existing data

Check if Instagram's interaction data (e.g., likes, saves, shares) and product schemas can support a similar feature. Conduct a power analysis to see if you can detect an effect of the expected size, and use propensity score matching to create comparable groups.

4. Quantify potential impact and address risks

Estimate the lift in engagement metrics from Facebook's results and project it onto Instagram's user base, accounting for differences in user behavior and product context. Identify potential cannibalization or negative effects and propose mitigation.

5. Present a compelling narrative to the PM

Synthesize findings into a clear story: the evidence from Facebook, the applicability to Instagram, the expected impact, and the recommended next steps (e.g., a limited A/B test). Highlight the low cost of testing and the potential upside.

Key Points to Mention

  • Causal inference methods: difference-in-differences, synthetic control, propensity score matching
  • Proxy metrics for engagement: CTR, session depth, time spent, saves/shares
  • Facebook as a natural experiment and external validity considerations
  • Power analysis and minimum detectable effect to ensure Instagram data can detect meaningful changes
  • Product schema alignment: ensuring similar product taxonomies and interaction types
  • Risk of cannibalization and negative user experience, and how to monitor

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

Q2

How would you design the experiment for this feature launch, and specifically how would you choose the randomization unit given potential network effects between users?

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

Network effects on Instagram are real and I knew user-level randomization would be contaminated fast.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the feature and the nature of potential network effects, then propose a randomization unit that balances bias reduction with practical constraints. Discuss trade-offs between user-level, cluster-level, and other randomization units, and suggest methods to detect and mitigate interference.

Pro tip: Mention that you would run a power analysis and consider variance reduction techniques like CUPED, and that you would pre-register the analysis plan to avoid p-hacking.

1. Clarify feature and network effects

Ask questions to understand how the feature works and how users interact. Identify if network effects are likely (e.g., social features) and their direction (positive or negative).

2. Choose randomization unit

Evaluate options: user-level, cluster-level (e.g., by social graph clusters), or time-based. Consider bias-variance trade-off and feasibility.

3. Design experiment and metrics

Define primary and guardrail metrics. Plan for sufficient power, and consider using variance reduction techniques. Decide on duration and sample size.

4. Address interference

If network effects exist, use cluster randomization or other methods like ego-cluster randomization. Consider measuring spillover effects.

5. Analyze and interpret

Use appropriate statistical methods (e.g., cluster-robust standard errors). Check for interference and adjust analysis if needed.

Key Points to Mention

  • Network effects can violate SUTVA and bias results.
  • Cluster randomization reduces interference but increases variance.
  • Trade-off between bias and variance when choosing randomization unit.
  • Use of graph clustering algorithms to form balanced clusters.
  • Power analysis and sample size calculation for cluster randomization.
  • Techniques like CUPED for variance reduction.

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

Q3

What would your primary and secondary metrics be for this experiment, keeping in mind engagement is the goal and not revenue? What guardrails would you set and what event logging would you add?

Product Analytics & MetricsA/B Testing & Experimentation
Author's notes

Primary metric I went with was click-through rate on the button plus downstream saves and session depth.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the experiment's goal and the specific engagement behavior you're targeting, then define a primary metric that directly measures that behavior and secondary metrics that capture broader engagement quality. Propose guardrails to ensure no harm to user experience or long-term value, and outline event logging needed to compute these metrics and diagnose results.

Pro tip: Tie your primary metric to a concrete user action that reflects meaningful engagement (e.g., daily active usage or content interactions), not just time spent, and explicitly state how you'd validate it with a counter-metric to avoid optimizing for vanity.

1. Clarify goal and scope

Restate the experiment's objective (engagement, not revenue) and identify the specific user behavior or segment you aim to influence. Ask clarifying questions if needed to align on what 'engagement' means for this product.

2. Define primary metric

Choose a single primary metric that directly measures the targeted engagement behavior (e.g., daily active users, sessions per user, or meaningful actions per user). Justify why it's the best proxy for the goal and how it will be measured.

3. Select secondary metrics

Pick 2-3 secondary metrics that capture other dimensions of engagement (e.g., depth, breadth, retention) to provide a fuller picture and support diagnosis if the primary metric moves.

4. Set guardrails

Identify guardrail metrics to monitor for unintended negative consequences, such as user satisfaction, performance, or long-term retention. Specify thresholds for alerting if guardrails degrade.

5. Plan event logging

List the events and properties needed to compute the metrics, including user actions, timestamps, and dimensions for segmentation. Ensure logging is consistent and can support both primary and guardrail analyses.

Key Points to Mention

  • Primary metric should be a direct measure of the engagement goal (e.g., DAU, actions per user) and not revenue.
  • Secondary metrics provide context and help explain why the primary metric moved (e.g., session duration, retention).
  • Guardrails protect against negative side effects (e.g., user reports, unsubscribes, performance degradation).
  • Event logging must capture the necessary user actions with sufficient granularity for segmentation and diagnosis.
  • Consider counter-metrics to avoid optimizing for quantity over quality of engagement.
  • Define success criteria and minimum detectable effect upfront to ensure the experiment is adequately powered.

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

Q4

Walk through how you'd compute sample size and minimum detectable effect under cluster randomization. What assumptions do you need and how would you handle multiple looks at the data during the experiment?

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

The cluster randomization adjustment to sample size tripped me up a bit.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by framing the problem: cluster randomization introduces intra-cluster correlation, so you need to adjust sample size using the design effect (1 + (m-1)ρ). Then explain how to compute MDE by inverting the power formula with the adjusted variance, and finally discuss sequential testing methods like alpha-spending or always-valid inference to handle multiple looks.

Pro tip: Emphasize that in practice, you often don't know ρ a priori, so you'd estimate it from historical data or run a pilot, and you might use a variance reduction technique like CUPED to improve sensitivity. Also, mention that multiple looks require pre-registration of the analysis plan to avoid p-hacking.

1. Define parameters and assumptions

Specify the outcome metric, cluster size (m), number of clusters (k), intra-cluster correlation (ρ), significance level (α), power (1-β), and minimum detectable effect (MDE) or sample size. Assume clusters are independent and outcomes within clusters are exchangeable.

2. Adjust sample size for clustering

Compute the design effect (DE = 1 + (m-1)ρ) and inflate the standard sample size formula by DE. For continuous outcomes, n_individual = 2*(z_{α/2}+z_β)^2 * σ^2 / Δ^2, then n_clusters = n_individual * DE / m.

3. Compute MDE given sample size

Invert the power formula: MDE = (z_{α/2}+z_β) * sqrt( (σ^2 * DE) / (k*m) ) for a two-sample t-test. For proportions, use the appropriate variance formula with DE.

4. Handle multiple looks

Use group sequential methods (e.g., O'Brien-Fleming alpha-spending) or always-valid confidence sequences to control type I error. Pre-specify the number and timing of interim analyses, and adjust α at each look.

5. Validate and iterate

Simulate the design to check power and false positive rates under realistic ρ and cluster size distributions. If ρ is uncertain, conduct sensitivity analysis or use a Bayesian approach to incorporate prior uncertainty.

Key Points to Mention

  • Intra-cluster correlation (ICC) and design effect formula
  • Sample size inflation for cluster randomization
  • MDE calculation by inverting power formula
  • Sequential testing methods: alpha-spending, group sequential, always-valid inference
  • Pre-registration of analysis plan to avoid p-hacking
  • Variance reduction techniques like CUPED or stratification

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

Q5

How would you define launch criteria combining statistical and practical significance, and how would you stage the rollout if results are promising in some segments but not others?

A/B Testing & ExperimentationProduct StrategyCross-functional Alignment
Author's notes

Said I'd require both p-value below threshold and a lift above some pre-agreed minimum meaningful effect, not just one or the other.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining launch criteria that balance statistical significance (e.g., p-value < 0.05) with practical significance (e.g., minimum detectable effect or business impact threshold). Then, outline a staged rollout plan that addresses heterogeneous treatment effects, using techniques like segment-level analysis and sequential testing to decide whether to launch fully, partially, or iterate.

Pro tip: Emphasize that practical significance should be tied to business metrics (e.g., revenue, engagement) and that segment-level results require correction for multiple comparisons to avoid false positives. Also, mention the importance of pre-registering the analysis plan to maintain rigor.

1. Define Success Metrics and Thresholds

Identify primary and secondary metrics, and set both statistical significance (alpha) and practical significance (minimum effect size) thresholds based on business impact.

2. Analyze Overall and Segment-Level Results

Conduct overall hypothesis tests and segment-level analyses (e.g., by user demographics, geography) while adjusting for multiple comparisons.

3. Evaluate Trade-offs and Risks

Assess whether segment differences are meaningful and consistent, and consider potential risks of partial rollout (e.g., user experience inconsistency, technical complexity).

4. Design Staged Rollout Plan

If results are promising in some segments, propose a phased rollout starting with those segments, with clear go/no-go criteria and monitoring for adverse effects.

5. Monitor and Iterate

Continuously monitor metrics post-launch, and be prepared to iterate or expand based on real-world performance and feedback.

Key Points to Mention

  • Statistical significance (p-value, confidence intervals) vs. practical significance (effect size, business impact)
  • Minimum Detectable Effect (MDE) and power analysis to determine sample size
  • Heterogeneous treatment effects and segment-level analysis
  • Multiple comparisons correction (e.g., Bonferroni, Benjamini-Hochberg) when analyzing segments
  • Staged rollout strategies: canary release, phased rollout, A/B testing in production
  • Cross-functional alignment with product, engineering, and business teams to define criteria and rollout plan

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