← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Intermediate

Intermediate
May 2026

Summary

Interviewed for a Data Scientist role at Meta. One question, stats-heavy, the kind where you either have the intuition baked in or you're fumbling through it in real time.

Questions Asked (1)

Q1

How would you describe the distribution of comment counts across Facebook users? Sketch or explain the shape and justify why you chose it.

Product Analytics & MetricsData Modeling
Author's notes

I knew it was going to be right-skewed but I second-guessed myself on whether to say log-normal or power-law.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the metric (comments per user over a given period) and the population (all Facebook users vs. active users). Then describe the expected shape as a heavily right-skewed distribution, likely following a power law or log-normal, and justify it by referencing user engagement patterns and platform dynamics.

Pro tip: Acknowledge that the distribution may vary by user segment (e.g., power users vs. casual users) and that the 'long tail' often contains bots or inactive accounts, which can affect the analysis. This shows you think about data quality and segmentation.

1. Clarify the metric and scope

Define what 'comment counts' means: comments made by a user, received by a user, or both? Specify the time frame (e.g., daily, monthly) and the population (all users, active users).

2. Describe the expected shape

State that the distribution is likely right-skewed, with a small number of users accounting for a large proportion of comments, and a long tail of users with few or zero comments.

3. Justify the shape with reasoning

Explain that engagement follows the Pareto principle (80/20 rule) due to varying user motivations, network effects, and content virality. Also mention that many users are passive consumers, leading to a spike at zero.

4. Consider alternative distributions and nuances

Discuss whether it might be log-normal (if multiplicative factors) or power law (if preferential attachment). Note that the shape could differ for comments received vs. made, and by user demographics.

5. Sketch or describe the curve

If sketching, draw a curve with a high peak near zero, rapidly declining, and a long tail to the right. Label axes: x-axis for number of comments, y-axis for frequency or density.

Key Points to Mention

  • Right-skewed distribution with a long tail
  • Pareto principle (80/20 rule) in user engagement
  • Presence of zero-comment users (passive consumers)
  • Potential power law or log-normal distribution
  • Differences between comments made vs. received
  • Impact of bots, spam, or inactive accounts on the tail

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