Design a A/B Testing for LinkedIn
Last updated: April 1, 2026
Quick Overview
Design a multi-tenant a/b testing system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.
LinkedIn
System Design
Product Manager
Product Manager
Technical Screen
System Design
Hard
464
12
4,454 solved
Design a multi-tenant a/b testing system that handles millions of requests. Discuss trade-offs in consistency, availability, and performance.
System design interviews at LinkedIn typically last 45-60 minutes. You are expected to drive the conversation, starting from requirements gathering through to a detailed architecture. The interviewer will evaluate your ability to handle ambiguity and make practical engineering decisions.
What the Interviewer Expects
- Drive the design discussion proactively with minimal interviewer guidance
- Perform detailed capacity estimation and use it to inform design decisions
- Design for global scale with multi-region deployment and data consistency
- Deep dive into 2-3 critical components with implementation-level detail
- Address complex trade-offs: CAP theorem, eventual consistency, conflict resolution
- Discuss operational excellence: deployment strategy, chaos engineering, SLOs/SLIs
- Propose a phased rollout plan from MVP to full-scale system
Key Topics to Cover
Security and authentication
Caching strategies (local, distributed, CDN)
Consistency models and replication
Database selection and data modeling
Message queues and async processing
Failure handling and fault tolerance
How to Approach This
- Start by clarifying functional and non-functional requirements with the interviewer.
- Estimate the scale: QPS, storage, bandwidth. This drives your design decisions.
- Draw a high-level architecture first, then deep dive into 1-2 critical components.
- Discuss trade-offs explicitly (e.g., consistency vs availability, SQL vs NoSQL).
- Address failure scenarios, monitoring, and how the system handles 10x traffic spikes.
Possible Follow-up Questions
- What would the deployment pipeline look like for this system?
- How would you handle a region-wide outage?
- What monitoring and alerting would you set up on day one?
Practice a Similar Problem on Codemia
Solve a related problem with our interactive workspace, get AI feedback, and view detailed solutions.
Solve on CodemiaSample Answer
Requirements
- Functional Requirements:
- User Segmentation: Ability to segment users based on demographics, activity levels, and interests.
- Experiment Creation: Users should be able to create ...
Capacity Estimation
- Estimating Load:
- Assume LinkedIn has approximately 900 million users.
- If 10% of users participate in A/B tests daily, that’s 90 million users.
- Assuming an average of 5 requests per u...
Submit Your Answer
Markdown supported