← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Intermediate

IntermediatePrefer not to say
Jun 2026Remote

Summary

Data Scientist interview at Meta centered almost entirely on a bad-account detection scenario. The questions ranged from basic probability to model evaluation to platform-level thinking, so you need to be comfortable jumping between math and product reasoning in the same conversation.

Questions Asked (10)

Q1

Given that 1% of accounts are bad and bad accounts send friend requests at 10x the rate of good accounts, what's the probability that a single received friend request came from a bad account?

Product Analytics & MetricsAlgorithms & Data Structures
Author's notes

This is a Bayes problem dressed up in product clothing.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Use Bayes' theorem to update the prior probability of a bad account given the evidence of a friend request. Define the base rates and likelihoods clearly, then compute the posterior probability.

Pro tip: State your assumptions explicitly (e.g., independence of friend requests, equal exposure) and mention that in practice you'd validate with A/B tests or holdout data. This shows rigor and business awareness.

1. Define the events and probabilities

Let B be the event that an account is bad, and F be the event that a friend request is received. Given P(B) = 0.01, so P(Good) = 0.99. Also, the rate of friend requests from bad accounts is 10 times that of good accounts.

2. Set up the likelihood ratio

Let r be the rate of friend requests from a good account. Then the rate from a bad account is 10r. So P(F|B) = 10r and P(F|Good) = r. The actual value of r cancels out in the calculation.

3. Apply Bayes' theorem

Compute P(B|F) = P(F|B)P(B) / [P(F|B)P(B) + P(F|Good)P(Good)] = (10r * 0.01) / (10r * 0.01 + r * 0.99) = 0.1 / (0.1 + 0.99) = 0.1 / 1.09 ≈ 0.0917.

4. Interpret the result

The probability is approximately 9.17%, meaning that despite the 10x higher rate, the low base rate of bad accounts (1%) keeps the posterior probability relatively low.

5. Discuss implications and assumptions

Mention that this assumes independence and that the rates are constant. In reality, you might need to consider other factors like account age, activity level, etc. Also, note that the result is sensitive to the base rate.

Key Points to Mention

  • Bayes' theorem and its application to base rate problems
  • The importance of the base rate (1% bad accounts) in the final probability
  • The likelihood ratio (10x) and how it affects the posterior
  • The assumption that the friend request rates are constant and independent of other factors
  • The cancellation of the unknown rate r in the calculation
  • The final probability is about 9.17%, which is less than 10% despite the 10x rate

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

Q2

Out of five friend requests received, what's the probability that at least one comes from a bad account?

Algorithms & Data StructuresProduct Analytics & Metrics
Author's notes

Complement rule, pretty mechanical once you have the per-request probability from the first part.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Clarify the assumptions: define what constitutes a 'bad account' and estimate its base rate from historical data. Then model the number of bad friend requests among five as a binomial random variable and compute the probability of at least one using the complement rule. Finally, discuss how the estimate might be refined with additional signals or Bayesian updating.

Pro tip: Show that you think like a data scientist by acknowledging that the base rate is rarely known with certainty and proposing a Bayesian approach or sensitivity analysis to account for uncertainty in the estimate.

1. Clarify definitions and assumptions

Define what a 'bad account' means (e.g., fake, spam, compromised) and assume a constant probability p for each friend request being from a bad account, independent of others.

2. Estimate the base rate p

Use historical data or domain knowledge to estimate the probability p that a randomly selected friend request comes from a bad account. If data is unavailable, propose a reasonable range or prior distribution.

3. Model the number of bad requests

Let X be the number of bad friend requests among 5. Assume X ~ Binomial(n=5, p). The probability of at least one is P(X ≥ 1) = 1 - P(X = 0) = 1 - (1-p)^5.

4. Compute and interpret

Plug in the estimated p to compute the probability. Discuss how the result changes with different p values and what it means for the product or user safety.

5. Discuss extensions and limitations

Mention that independence may not hold (e.g., coordinated attacks), and suggest using Bayesian methods or incorporating additional features to improve the estimate.

Key Points to Mention

  • Binomial distribution and independence assumption
  • Complement rule: P(at least one) = 1 - P(none)
  • Base rate estimation from historical data
  • Sensitivity analysis for different p values
  • Bayesian updating with prior knowledge
  • Limitations: non-independence, varying p across requests

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

Q3

Your classifier has a 95% true positive rate and 95% true negative rate. If it flags an account as bad, what's the actual probability that account is truly malicious?

Product Analytics & MetricsAlgorithms & Data Structures
Author's notes

Classic low base rate trap, and I still fumbled it for a second.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Recognize this as a base rate problem and apply Bayes' theorem. Explain that the answer depends on the prevalence of malicious accounts, which is not given, so you must state the assumption or ask for it. Then compute the posterior probability using the provided rates and a reasonable base rate.

Pro tip: Always clarify the base rate before calculating; in practice, malicious accounts are rare, so even with high accuracy, the probability can be surprisingly low. Mention that this is why precision matters more than accuracy in imbalanced settings.

1. Identify the problem type

Recognize that this is a conditional probability question requiring Bayes' theorem, not just the given true positive and true negative rates.

2. Define the events and knowns

Let M be the event that an account is malicious, and F be the event that it is flagged. Given: P(F|M)=0.95, P(not F|not M)=0.95, so P(F|not M)=0.05. The unknown is P(M|F).

3. Introduce the base rate

State that P(M), the prevalence of malicious accounts, is needed. If not provided, assume a realistic value (e.g., 1%) or ask the interviewer for it.

4. Apply Bayes' theorem

Compute P(M|F) = P(F|M)P(M) / [P(F|M)P(M) + P(F|not M)P(not M)]. Plug in the numbers and calculate.

5. Interpret and discuss implications

Explain the result in context: even with 95% accuracy, if the base rate is low, the probability that a flagged account is truly malicious can be low. Discuss the importance of precision and potential business impact.

Key Points to Mention

  • Bayes' theorem and conditional probability
  • Base rate (prevalence) of malicious accounts
  • True positive rate (sensitivity) and true negative rate (specificity)
  • False positive rate and its impact on precision
  • The confusion matrix and derived metrics like precision and recall
  • The effect of class imbalance on classifier performance

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

Q4

What types of data and features would you use to build a classifier that distinguishes bad accounts from good ones?

Data ModelingProduct Analytics & Metrics
Author's notes

Talked about behavioral signals like friend request volume, message patterns, account age, login locations, and whether the account had been reported.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the business context and definition of 'bad' accounts, then outline a structured approach covering data sources, feature engineering, modeling, and evaluation. Emphasize the importance of aligning with product goals and handling class imbalance.

Pro tip: Mention that you would collaborate with cross-functional teams to define 'bad' and incorporate feedback loops, showing product sense and stakeholder awareness. Also, highlight the need for model interpretability to explain decisions to trust and safety teams.

1. Define 'Bad' Accounts

Clarify what constitutes a bad account (e.g., spam, fake, compromised) and align with business objectives. Consider both rule-based and ML-based definitions.

2. Identify Data Sources

List relevant data: account metadata, user behavior logs, content, network connections, and external signals. Ensure data quality and coverage.

3. Engineer Features

Create features from raw data: behavioral (login frequency, session duration), content (text, images), network (friend requests, interactions), and temporal patterns.

4. Select and Train Model

Choose appropriate algorithms (e.g., gradient boosting, neural networks) considering class imbalance, scalability, and interpretability. Use techniques like resampling or anomaly detection.

5. Evaluate and Iterate

Define metrics (precision, recall, AUC) aligned with business costs. Validate with holdout sets, monitor drift, and incorporate human feedback for continuous improvement.

Key Points to Mention

  • Account metadata: age, verification status, profile completeness
  • Behavioral features: login patterns, posting frequency, interaction rates
  • Content-based features: text analysis, image hashing, spam keywords
  • Network features: graph-based metrics like centrality, clustering coefficient
  • Temporal features: time since creation, activity bursts, diurnal patterns
  • Handling class imbalance: techniques like SMOTE, class weights, or anomaly detection

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

Q5

How would you estimate whether the bad account problem is large enough to actually do something about it?

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

Proposed random sampling plus stratified sampling to get a reliable prevalence estimate.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining what a 'bad account' is and quantifying its prevalence and impact on key metrics. Then, assess whether the problem is large enough to warrant action by comparing its impact to other priorities and estimating the potential ROI of a solution.

Pro tip: Frame the problem in terms of business impact and opportunity cost—show that you can prioritize based on data, not just intuition. Quantify the potential lift from fixing the issue and compare it to the effort required.

1. Define and Measure the Problem

Clearly define what constitutes a 'bad account' (e.g., fraudulent, inactive, low-quality) and quantify its prevalence using available data. Calculate the proportion of bad accounts and their impact on key metrics like revenue, engagement, or retention.

2. Assess Impact on Business Metrics

Estimate how much bad accounts affect critical business metrics. For example, if bad accounts represent 5% of users but cause 20% of support tickets, the impact is disproportionate. Use cohort analysis or regression to isolate the effect.

3. Estimate Potential Improvement

Model the potential improvement if the problem were solved. For instance, if you could reduce bad accounts by 50%, what would be the expected lift in revenue or reduction in costs? Use historical data or A/B tests to estimate effect sizes.

4. Compare to Other Priorities

Evaluate the opportunity cost: is this problem more impactful than other initiatives? Consider the effort required (engineering, data science, product) and the expected ROI. Use a prioritization framework like RICE or ICE.

5. Recommend Action or Further Investigation

Based on the analysis, recommend whether to proceed with a solution, run a pilot, or gather more data. If the problem is large enough, propose an experiment to validate the impact of a potential fix.

Key Points to Mention

  • Define 'bad account' clearly and align with stakeholders on the definition.
  • Quantify prevalence and impact using data (e.g., percentage of accounts, revenue at risk).
  • Use metrics like revenue, engagement, retention, and support costs to measure impact.
  • Estimate potential lift from solving the problem and compare to effort (ROI).
  • Consider opportunity cost and prioritization frameworks (e.g., RICE).
  • Propose an A/B test or pilot to validate the solution's effectiveness.

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

Q6

How would you define what makes an account 'bad' in the first place?

Product Sense & IdeationData Modeling
Author's notes

Spammers, bots, scammers, coordinated inauthentic behavior.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying that 'bad' is context-dependent and must be tied to a specific business objective, such as reducing harm or improving user experience. Then propose a framework that combines product policy definitions with data-driven signals, and emphasize the importance of validation and iteration.

Pro tip: Acknowledge that 'bad' is not a fixed label but a dynamic, multi-dimensional construct that evolves with product changes and adversarial behavior. Show you can balance precision and recall trade-offs based on the cost of false positives vs. false negatives.

1. Clarify the Objective

Ask what problem we're solving: is it reducing spam, fake accounts, or harmful content? Align with stakeholders on the business goal and the cost of errors.

2. Define 'Bad' Operationally

Translate the objective into measurable criteria, such as policy violations, low-quality interactions, or anomalous behavior. Consider both rule-based and ML-based definitions.

3. Identify Data Signals

List potential signals from user behavior, content, network, and metadata. Prioritize signals that are predictive, scalable, and robust to adversarial manipulation.

4. Validate and Iterate

Use labeled data and experiments to test definitions. Measure precision, recall, and business impact, and refine as needed.

5. Monitor and Adapt

Set up ongoing monitoring to detect drift and new bad behaviors. Update definitions as the product and adversaries evolve.

Key Points to Mention

  • The definition of 'bad' must be tied to a specific business objective and user harm.
  • Consider multiple dimensions: policy violations, low-quality content, fake engagement, etc.
  • Use a combination of rule-based heuristics and machine learning models.
  • Evaluate trade-offs between false positives and false negatives based on business costs.
  • Leverage both labeled data and unsupervised anomaly detection.
  • Plan for adversarial adaptation and concept drift over time.

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

Q7

How does a significant presence of malicious users affect the platform's health and reputation?

Product StrategyProduct Sense & Ideation
Author's notes

More of a product thinking question.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by defining what constitutes a 'significant presence' of malicious users and the dimensions of platform health (e.g., user engagement, trust, safety) and reputation (e.g., brand perception, media coverage). Then, analyze the direct and indirect effects, using a structured framework to cover user, content, and business impacts, and propose metrics to quantify these effects.

Pro tip: Quantify the impact with metrics like DAU/MAU changes, trust scores, and sentiment analysis, and suggest potential mitigation strategies to show proactive thinking. This demonstrates a data-driven and strategic mindset, which is crucial for a Data Scientist at Meta.

1. Define Malicious Users and Platform Health

Clarify what you mean by malicious users (e.g., spammers, scammers, harassers) and define platform health (e.g., user engagement, retention, well-being) and reputation (e.g., public perception, brand trust).

2. Identify Direct Impacts

Discuss how malicious users directly affect platform health, such as degrading content quality, increasing moderation costs, and causing user churn. Also, consider direct reputation hits like negative press and user complaints.

3. Analyze Indirect and Long-term Effects

Examine indirect effects like erosion of user trust, reduced engagement from legitimate users, and potential regulatory scrutiny. Long-term reputation damage can lead to difficulty in attracting new users and advertisers.

4. Quantify with Metrics

Propose metrics to measure the impact, such as changes in DAU/MAU, NPS, sentiment analysis of user feedback, and ad revenue trends. This shows a data-driven approach.

5. Suggest Mitigation and Trade-offs

Briefly mention potential strategies to mitigate the impact (e.g., AI moderation, community guidelines) and discuss trade-offs (e.g., false positives, privacy concerns).

Key Points to Mention

  • User trust and safety as a core platform value
  • Impact on engagement metrics (e.g., DAU, session time)
  • Reputation damage through media and social sentiment
  • Increased moderation and operational costs
  • Regulatory and legal risks (e.g., GDPR, FTC)
  • Advertiser confidence and revenue implications

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

Q8

What are the downstream effects on real users when bad accounts send friend requests at high volume?

Product Analytics & MetricsProduct Sense & Ideation
Author's notes

Short answer: people get annoyed, start ignoring requests, and eventually trust the platform less.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by mapping the user journey when receiving a friend request, then identify potential harms at each stage (notification, decision, post-acceptance). Quantify impact using metrics like spam report rate, request rejection rate, and user engagement changes, and propose mitigation strategies.

Pro tip: Frame your answer around Meta's key metrics (e.g., DAU, meaningful social interactions) and show you understand the trade-off between blocking bad actors and minimizing false positives that could harm legitimate users.

1. Map the user journey

Outline the steps a user takes when receiving a friend request: notification, viewing request, deciding to accept/ignore, and post-acceptance interactions.

2. Identify potential harms

For each step, list possible negative effects: notification overload, decision fatigue, privacy concerns, unwanted content exposure, and reduced trust.

3. Quantify with metrics

Propose metrics to measure each harm, such as request rejection rate, spam report rate, time spent on requests, and changes in user engagement.

4. Consider second-order effects

Discuss how these harms might cascade: users becoming less active, sharing less content, or leaving the platform, impacting network health.

5. Propose mitigations and trade-offs

Suggest interventions like rate limiting, ML-based detection, and user controls, and discuss trade-offs between blocking bad actors and false positives.

Key Points to Mention

  • Notification fatigue and increased cognitive load from high volume of requests
  • Erosion of trust in the platform and friend request system
  • Potential exposure to spam, scams, or malicious content after acceptance
  • Decrease in meaningful social interactions and user engagement
  • Increased load on support and reporting systems
  • Trade-offs between aggressive filtering and false positives that may block legitimate requests

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

Q9

How would you think about adjusting the classification threshold, and what are the tradeoffs between false positives and false negatives here?

Technical Trade-offsA/B Testing & Experimentation
Author's notes

This is where the cost-benefit framing matters.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by tying the threshold to the product's objective and the relative cost of each error type, then propose a data-driven method to select it (e.g., cost-sensitive optimization or precision-recall curve analysis). Acknowledge that the optimal threshold depends on the specific use case and may require experimentation to validate business impact.

Pro tip: Mention that you would validate the chosen threshold through an online A/B test, because offline metrics don't always capture user behavior and long-term effects. Also, consider that the threshold might need to be personalized or dynamic based on context.

1. Clarify the business objective and error costs

Ask or infer what the model is used for and quantify the cost of a false positive versus a false negative in that context. For example, in content moderation, a false negative (harmful content shown) may be costlier than a false positive (over-removal).

2. Analyze the precision-recall tradeoff

Plot the precision-recall curve and examine how precision and recall change with threshold. Identify the threshold range that aligns with the business constraints (e.g., maintain precision above 90%).

3. Choose an optimization criterion

Select a metric that reflects the business goal, such as expected cost, F-beta score, or profit. Optimize the threshold to maximize this metric on validation data, possibly using cost-sensitive learning.

4. Validate offline and plan online experiment

Evaluate the chosen threshold on a holdout set and simulate business impact. Then design an A/B test to compare the new threshold against the current one, measuring both primary and guardrail metrics.

5. Monitor and adjust over time

After deployment, continuously monitor performance and re-evaluate the threshold as data distributions or business costs change. Consider dynamic thresholding if appropriate.

Key Points to Mention

  • Cost-sensitive evaluation: assign monetary or utility values to FP and FN.
  • Precision-recall curve and ROC curve: understand tradeoffs and choose based on business needs.
  • F-beta score: adjust beta to weight recall more than precision (or vice versa).
  • A/B testing: validate threshold changes online to capture real user impact.
  • Business metrics alignment: ensure the threshold optimizes the ultimate KPI (e.g., user engagement, revenue).
  • Class imbalance: threshold adjustment is often more effective than resampling for imbalanced data.

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

Q10

How would you scale this detection system fairly across billions of accounts without creating disproportionate harm for certain user groups?

System DesignTechnical Trade-offs
Author's notes

Fairness across demographics is genuinely hard and I didn't have a crisp answer.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the system's goals and constraints, then propose a scalable architecture that includes fairness-aware components. Discuss trade-offs between detection accuracy, fairness, and computational cost, and outline how you would measure and mitigate disproportionate harm across user groups.

Pro tip: Emphasize the importance of defining fairness metrics upfront and involving cross-functional teams (e.g., policy, legal) to align on acceptable trade-offs. Also, mention the need for continuous monitoring and auditing to catch emerging biases.

1. Clarify Objectives and Constraints

Ask questions to understand what 'fairness' means in this context, what the detection system aims to detect, and any regulatory or business constraints. This ensures your answer is aligned with the interviewer's expectations.

2. Propose a Scalable Architecture

Outline a distributed system design that can handle billions of accounts, such as using sharding, stream processing, and efficient storage. Highlight how fairness considerations can be integrated at each layer.

3. Define Fairness Metrics and Monitoring

Specify quantitative fairness metrics (e.g., demographic parity, equalized odds) and describe how you would monitor them in production. Discuss the need for ground truth labels and potential biases in data.

4. Mitigate Disproportionate Harm

Explain techniques to reduce bias, such as reweighting training data, adversarial debiasing, or post-processing adjustments. Discuss trade-offs between fairness and detection performance.

5. Iterate and Audit

Describe a process for continuous evaluation, including A/B testing, fairness audits, and stakeholder feedback. Emphasize the need for transparency and accountability.

Key Points to Mention

  • Scalability techniques: distributed computing, sharding, stream processing
  • Fairness definitions and metrics: demographic parity, equalized odds, disparate impact
  • Bias mitigation methods: pre-processing, in-processing, post-processing
  • Trade-offs: accuracy vs. fairness, latency vs. thoroughness, cost vs. benefit
  • Monitoring and auditing: dashboards, alerts, regular fairness reviews
  • Cross-functional collaboration: policy, legal, product teams

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