← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Senior

Senior
May 2026

Summary

Meta DS interview, one meaty experimentation question about hashtag recommendations. The whole session was basically a deep dive into experiment design and I felt underprepared for how many sub-topics they expected you to cover.

Questions Asked (1)

Q1

How would you design and run an experiment to evaluate a new hashtag recommendation algorithm, and how would you decide on test versus control group assignment?

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

I started with randomization unit and felt okay there, user-level made sense to avoid exposure bleed.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining clear success metrics (e.g., CTR, engagement, diversity) and then outline a randomized controlled experiment with proper randomization unit and sample size calculation. Emphasize the importance of avoiding network effects and ensuring valid inference through techniques like cluster randomization or switchback testing.

Pro tip: At Meta, where social influence is strong, consider using cluster randomization by user or by social graph to mitigate interference, and always run an A/A test beforehand to validate your experimentation setup.

1. Define Hypothesis and Metrics

Clearly state the hypothesis (e.g., new algorithm increases engagement) and select primary and guardrail metrics (e.g., CTR, time spent, diversity, user satisfaction).

2. Choose Randomization Unit and Design

Decide whether to randomize at user, session, or cluster level (e.g., by social graph) to avoid interference; consider switchback or geo-based randomization if needed.

3. Calculate Sample Size and Power

Determine required sample size based on expected effect size, power (80%), significance level (5%), and variance; account for network effects if present.

4. Run Experiment and Monitor

Launch the experiment, monitor for SRM (sample ratio mismatch), and ensure data quality; avoid peeking or early stopping unless using sequential testing.

5. Analyze Results and Decide

Analyze primary and guardrail metrics, check for heterogeneous treatment effects, and make a ship/no-ship decision based on statistical and practical significance.

Key Points to Mention

  • Randomization unit (user vs. session vs. cluster) and its implications for interference
  • Network effects and solutions like cluster randomization or switchback testing
  • Sample size calculation and power analysis
  • Guardrail metrics to ensure no negative impact (e.g., user satisfaction, diversity)
  • A/A testing to validate experiment setup
  • Heterogeneous treatment effects and subgroup analysis

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