Meta Interview Questions — Page 234

Meta·Software Engineer·Technical Phone Screen

Jul 2026

SQL-heavy Data Engineer screen at Meta, four questions all centered on a library database schema. Nothing behavioral, just back-to-back query writing with some tricky aggregation logic thrown in.

Given a library database, write a SQL query to count books that are currently checked out and in good condition.
Among the checked-out, good-condition books from the previous query, what percentage have been renewed more than twice? Use that same subset as the denominator.
Return the top 3 book titles (or IDs) that have more than 10 copies and the longest cumulative lending time summed across all their loans.
Find the member and referrer pair with the largest difference in the number of reservations each has made, and return that pair along with the difference.

“Warm-up question, felt fine.”

View Post

Meta·Software Engineer·Onsite - System Design / Architecture

Jul 2026

System design round at Meta for a software engineer role. The problem was building an ad metrics backend from scratch, covering ingestion, aggregation, and query support at scale. Pretty meaty question with a lot of moving parts.

Design a backend service that ingests high-volume ad impression and click events and serves both near-real-time and historical aggregated metrics, with support for duplicate and late-arriving events.

“I started with the data model and event schema which felt safe, but I spent too long there and had to rush through the streaming pipeline.”

View Post

Meta·Software Engineer·Technical Phone Screen

Jul 2026

Meta SWE coding round, one problem the whole time: merge M sorted arrays and return the first K elements. Pretty standard heap question but the constraints pushed you toward thinking carefully about efficiency.

Given M sorted integer sequences, output the first K elements of the combined sorted order across all sequences, keeping duplicates.

“Min-heap is the move here.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

A Meta DS interview that went deep into causal inference and network experimentation, specifically around testing a misinformation warning label. The question was layered in a way I wasn't fully prepared for, covering everything from graph partitioning to adversarial detection.

You're testing a warning label shown before users reshare a suspected misinformation link. Since behavior spreads through a social graph, SUTVA is violated. Walk through how you'd design an experiment to measure total, direct, and spillover effects. Cover cluster construction, exposure mapping, outcome metrics, estimation strategy, power considerations, sensitivity analyses for contamination, and how you'd detect adversarial adaptation during the test.

“This one hit me like five questions stacked inside one.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Probability and statistics question for a Data Scientist role at Meta. Pretty straightforward Bayesian/independence stuff but the third question is a small gotcha if you're not paying attention.

If a chatbot response is considered 'good' only when it is both helpful and honest, and P(Helpful) = 0.8 and P(Honest) = 0.9 with the two being independent, what is P(Good) for a single response?
Using the same setup, what is the probability of getting two good responses in a row?
Given that the first three responses are all good, what is the probability that the fourth response is also good?

“Multiplication of independent probabilities.”

View Post

Meta·Software Engineer·Onsite - Product Sense / Strategy

Jul 2026

Went through a product growth case loop for a Product Analyst role at Meta. Four big case questions, all focused on Instagram and Facebook products. Heavy on metrics, funnels, and experimentation design. Left feeling like I could've gone deeper on the guardrails side of things.

How would you improve the Instagram Stories viewing experience, and how would you test one of your product ideas?
Boosted Posts MAU is your target metric. How do you assess whether doubling it in six months is realistic, and what levers would you pull across the user lifecycle?
How would you define an active Reels creator, diagnose supply-side bottlenecks in the creator funnel, and prioritize interventions for durable creator growth?
Logged-in user count on mobile Facebook is declining. How do you figure out if it's a measurement issue, a bug, a security tradeoff, or actual user behavior change?

“I jumped straight to ideas before really anchoring on a metric, which I think hurt me.”

View Post

Meta·Data Scientist·Technical Phone Screen

Jul 2026

Meta DS interview with a pretty brutal SQL question covering anomaly investigation across multiple tables. The whole thing felt like a take-home that somehow ended up in a live session, lots of moving parts and edge cases they clearly expected you to handle without prompting.

For each country and post type on a given date, compute Likes-per-DAU and its percent change versus the median value for the same weekday over the prior 8 weeks, excluding dates with an outage flag. Use window functions like PERCENTILE_CONT and pull DAU from the DailyActiveUsers table rather than counting users yourself.
Flag the top 3 countries with the largest percentage declines (at or below negative 10%) in Likes-per-DAU, and for each, break down how much of that decline comes from new users versus existing users. Return country, post type, percent change, and the share of decline attributable to new users.
For the flagged countries, compute the friend request acceptance rate over the most recent 14-day window and compare it to the prior 14-day window. Return the absolute and relative change, and do this with window functions rather than correlated subqueries.
Walk through a pandas approach that mirrors the SQL logic above, covering groupby, merge, rolling or expanding windows, and quantile operations.

“This is where I spent most of my mental energy.”

View Post

Meta·Software Engineer·Technical Phone Screen

Jul 2026

SQL-heavy technical screen for a Data Engineer role at Meta. One question, pretty deep on window functions and filtering logic. Felt okay during but second-guessed my tie-breaking clause afterward.

Given a copies table (copy_id, book_id) and a checkouts table (copy_id, checkout_date, return_date), find the top 3 book_ids that have more than 10 distinct copies and rank them by total borrowed duration (sum of day-level date differences across completed checkouts only), with ties broken by book_id ascending.

“The NULL filter tripped me up first.”

View Post

Meta·Software Engineer·Technical Phone Screen

Jul 2026

Phone screen for a senior SWE role at Meta, two coding problems back to back. The second one was this prefix-sum modulo thing that looks easy until you start handling edge cases and suddenly you're explaining language-specific modulo behavior to a skeptical interviewer.

Given an integer array and an integer k, determine whether there exists a contiguous subarray of length at least 2 whose sum is a multiple of k.

“The core trick is prefix sums mod k, store the first time you see each remainder and check if you've seen it before with enough distance between indices.”

View Post

Meta·Data Scientist·Onsite - System Design / Architecture

Jul 2026

Brutal system design question for a DS role at Meta, all centered around Instagram Shopping's ranking system. The scope was massive and I felt like I was playing catch-up the whole time.

Design a multi-objective ranking framework for Instagram Shopping's home feed that balances buyer GMV, buyer satisfaction, and seller fairness. Walk through your model architecture, how you aggregate competing objectives, and how you handle training data bias from historical rankings.
How would you enforce minimum exposure guarantees for sellers, including cold-start sellers with no historical data, without tanking overall feed quality?
How do you make the ranking system robust to fraudulent sellers and fake engagement signals?
The ranking system needs to support on-device prefetching within a 20 ms latency budget. How do you design for that constraint?
After launch, how do you monitor for fairness drift and fraud drift, and how do you throttle supply during demand spikes like flash sales?

“This is the kind of question where you think you know where to start and then realize five minutes in that you've already painted yourself into a corner.”

View Post