Top System Design Questions at Big Tech (2025)
The most frequently asked system design questions across Google, Meta, Amazon, Microsoft, and top unicorns — with a framework for approaching each one.
How to Approach Any System Design Interview
Before diving into specific questions, the most valuable thing you can do is develop a repeatable framework. Great system design interviews follow a predictable arc: clarify requirements and scale, define the API surface, sketch a high-level architecture, dive deep on the hard parts, and discuss tradeoffs and failure modes. Interviewers are not looking for a single correct answer — they're evaluating your ability to think systematically under ambiguity and communicate tradeoffs clearly.
A reliable starting sequence: ask about scale (QPS, users, data size), identify read-vs-write ratio, decide on data model, propose a simple architecture first, then iterate into a scalable version. Never skip directly to distributed solutions — show that you understand the simple case before complicating it.
The Most Frequently Asked System Design Questions
Based on aggregated reports from thousands of interview experiences, these are the questions that appear most often across top tech companies:
1. Design a URL Shortener (like bit.ly)
Tests hashing, database choice, redirect performance, and analytics. Key decisions: hash function, collision handling, custom aliases, expiration, read-heavy optimization.
2. Design Twitter / a Social Feed
One of the most common Meta and Twitter questions. Core challenge: fan-out on write vs fan-out on read for news feed. Introduces celebrity problem, timeline generation, and notification systems.
3. Design a Rate Limiter
Ubiquitous at Stripe, Cloudflare, and API-heavy companies. Tests understanding of token bucket vs leaky bucket vs sliding window algorithms, distributed rate limiting with Redis.
4. Design a Distributed Cache (like Redis/Memcached)
Fundamental infrastructure design. Covers cache eviction policies (LRU/LFU), consistent hashing for sharding, replication, and cache invalidation (the hard problem).
5. Design a Notification System
Common at Amazon and Meta. Multi-channel (push, email, SMS), fan-out at scale, reliability guarantees, user preference management, rate limiting per user.
6. Design a Search Autocomplete
Asked frequently at Google and LinkedIn. Core challenge: Trie vs prefix hash, top-k results, personalization, latency under 100ms requirements.
7. Design Google Drive / Dropbox
File storage, chunking, deduplication, sync protocols, conflict resolution. Asked at Google, Dropbox, and Microsoft.
8. Design a Ride-Sharing Service (like Uber)
Geospatial indexing (quadtree, geohash), real-time driver matching, surge pricing, map routing — popular at Uber, Lyft, and for any GPS/geolocation role.
9. Design a Video Streaming Platform (like YouTube/Netflix)
Encoding pipelines, CDN distribution, adaptive bitrate streaming. Common at Netflix, YouTube, and TikTok.
10. Design a Messaging System (like WhatsApp/Slack)
Message ordering, delivery guarantees, read receipts, end-to-end encryption basics, group messaging at scale.
Core Concepts You Must Know Cold
Every system design question ultimately recombines a relatively small set of building blocks. Understanding each deeply — including the tradeoffs between alternatives — prepares you for any variant you'll encounter.
Company-Specific Patterns
Different companies have different system design preferences. Google tends toward infrastructure and distributed systems questions. Meta favors social graph and feed-generation problems. Amazon often asks questions tied to AWS services (design a feature on top of S3, SQS, etc.). Fintech companies (Stripe, Robinhood, Jane Street) focus heavily on correctness, transactions, and idempotency.
Tailor your prep accordingly. If you're interviewing at Stripe, become fluent in payment idempotency, ledger systems, and distributed locking. If you're targeting Meta, master fan-out architectures and graph traversal at scale. Real interview reports filtered by company give you this signal — and it's the fastest way to calibrate.
See Real System Design Questions by Company
Filter by role, round, and seniority to find the exact system design questions asked at your target company.