Aim for 150-200 problems with a 70/30 split favoring medium and hard. Criteo's coding rounds often involve medium-level array/string manipulation and dynamic programming, but they frequently include one hard problem that tests optimization and clean code. Focus heavily on problems tagged 'Criteo' on LeetCode to understand their exact style.

The Bar Raiser is a deep-dive into Criteo's 12 Leadership Principles (like 'Customer Obsession' and 'Learn and Be Curious'). You must prepare concise, metric-driven stories using the STAR method that explicitly map to these principles. unlike a casual chat, this round is highly structured, and interviewers take detailed notes to assess cultural fit against a calibrated bar.

For SDE-2, expect to design a scalable subsystem (e.g., a real-time bidding component) with focus on APIs, data models, and trade-offs. For SDE-3, you'll lead a high-level architecture discussion for an end-to-end system (e.g., a new ad product), covering scalability, cost, team organization, and long-term tech debt. SDE-3 requires more depth on trade-off justification and cross-team impact.

The biggest mistake is jumping into code without clarifying requirements and edge cases. Criteo interviewers highly value collaborative problem-solving. You must verbally walk through your approach, discuss time/space complexity, and ask clarifying questions about constraints (e.g., data volume, latency requirements) before writing a single line of code.

The process usually takes 4-6 weeks. After the initial recruiter screen (1 week), you'll have 3-4 technical rounds (coding, system design, Bar Raiser) scheduled over 1-2 weeks. Team matching and offer approval can add 1-2 more weeks. Delays often occur in the team-matching phase for SDE-2/3 roles, where multiple teams may be considered.

Stand out by connecting your technical solutions to Criteo's business context. Mention how your design impacts ad performance, latency, or revenue. In the Bar Raiser, use data-rich stories with specific metrics (e.g., 'improved system throughput by 30%'). Showing genuine curiosity about their stack (Hadoop, Spark, real-time bidding) by asking insightful follow-up questions is critical.

Study the Criteo Engineering Blog and their tech talks on YouTube for concrete projects. Follow Criteo R&D on LinkedIn. For culture, deeply review their 12 Leadership Principles on their careers page. Use LeetCode's company-specific filter and Blind/Reddit threads for recent interview experiences, as their focus areas can shift quarterly.

Emphasize experience with large-scale data processing, performance optimization, and distributed systems. Use metrics in every bullet point (e.g., 'reduced pipeline latency by X ms'). In your initial calls, explicitly mention familiarity with Criteo's core domains—real-time bidding, machine learning for advertising, and cloud infrastructure—and express interest in their impact-driven, collaborative Paris/R&D environment.