Below the Surface: Real SQL Interview Questions from Meta and How I Solved Them

Meta Business Intelligence Analyst Interview Experience:

CTC - 40 LPA

SQL Questions Asked in Meta: A Practical Walkthrough for the Curious Analyst

There is a moment every analytics learner remembers—when raw SQL questions suddenly stop looking like puzzles from another planet, and begin to feel like stories with characters, conflicts, and neat resolutions. The Meta SQL interview is exactly that kind of scene. When I first encountered these questions, they didn’t feel like plain technical problems. They felt like real behaviour unfolding inside a massive social network, waiting to be decoded.

In this post, I want to take you on that same path. Not with heavy jargon or textbook style, but in a way that makes you pause, think, and then smile with a quiet haan bhai, samajh aaya moment. If you’re preparing for SQL interviews, building your analytical thinking, or exploring data-driven decision-making, these questions will sharpen your instincts.

Let’s walk through them one by one.

1. The Case of the “Average Post Hiatus”

Imagine millions of Facebook users posting their thoughts throughout 2024. Some share daily life moments, some disappear for weeks, some write only when inspiration strikes. If you trace each user’s first post of the year and their last, you get a personal storyline. The question is simple:
How many days lie between those two scenes for users who posted at least twice in 2024?

It looks like this:

posts table
user_id, post_id, post_date, post_content

What the interviewer wants is a blend of date logic, careful grouping, and filtering users who truly belong to 2024’s “active” league. These queries often test whether you think in timelines rather than just rows. Once you learn to read data like time-based chapters, your SQL becomes sharper.

2. Facebook Power Users: The People Who Never Log Off

Every platform has them—the super-active crowd. The ones who post multiple times a day and attract reactions like iron filings to a magnet. Meta defines a “power user” here as someone who posts at least twice daily and averages 150 comments or reactions per post.

Two tables drive this story:

user_post – user_id, post_id, post_date
post_interactions – post_id, comments, reactions

This question checks your ability to join behaviour and engagement. Platforms like Meta care deeply about users who keep the ecosystem alive. Understanding this logic helps you think like a product analyst, not just a query writer. As the saying goes, where there’s smoke, there’s fire, and where there’s high engagement, there’s usually impactful user behaviour to study.

3. Active User Retention: The Silent Pulse of July 2022

User retention is the oxygen of any digital platform. Without repeat activity, even a giant feels small. Here the question is to calculate Monthly Active Users (MAU) for July 2022, but only those who performed actions in both June and July.

The table is:

user_actions – user_id, event_id, event_type, event_date

To solve it, you need to think like someone who understands patterns. SQL becomes more meaningful when you start treating events as signals, not just entries. Retained users are the ones who come back, again and again, like old friends who never forget to call.

4. Friend Recommendations: When Data Plays Cupid

Meta wants to suggest new friends to users who show interest in two or more of the same private events. If two strangers keep signing up for the same parties, well, “birds of the same feather flock together”—data agrees too.

Here are the tables making this social magic possible:

friendship_status – user_a_id, user_b_id, status
event_rsvp – user_id, event_id, event_type, attendance_status, event_date

The trick is to filter only private events and consider users with attendance statuses like going or maybe. Then you identify pairs, avoid duplicates, and output unidirectional recommendations. These types of queries teach you the gentle art of handling self-joins and large social graphs. A little like arranging chairs at a shaadi—every pairing needs careful thought.

5. Average Number of Shares per Post: The Echo of a Story

When a user shares a post, they’re forwarding a piece of themselves. Measuring shares per user helps platforms understand whose voice amplifies far beyond their profile.

Tables:

user_posts – post_id, user_id, post_text, post_date
post_shares – share_id, post_id, share_date

This problem tests aggregation, grouping, and understanding connection between user-level activity and post-level behaviour. When you solve enough such cases, you start spotting the beating heart of user engagement.

Closing Thoughts

SQL questions like these don’t just measure technical skill. They measure how comfortably you move inside data—whether you think like an analyst who sees patterns behind the curtain. Working through real interview problems is one of the strongest ways to grow your analytical muscles. As you practice more scenarios like these, you’ll find that interviews, freelance projects, and real-world dashboards all start speaking in the same language.

If this kind of storytelling approach to analytics helps you learn better, the next steps become exciting. There is always another dataset waiting, another query asking to be cracked, another business problem ready to be decoded.

Data keeps the curious mind alive.