If you've spent months grinding LeetCode randomly, you've probably noticed something frustrating: the questions you practiced don't match what companies actually ask in interviews.
This guide is built on data from 10,385 real interview questions across 259 companies. You'll learn exactly what each top company asks, what patterns they love, and how to build a targeted 30–60 day prep plan—so you stop wasting time and start interviewing smarter.
Why Company-Wise Prep Matters
Here's the truth most people learn too late:
- Amazon loves trees and design problems — Amazon Leadership Principles show up even in code
- Google favors graphs, DP, and optimization — they want to see you think at scale
- Meta focuses on practical algorithms — social network = graph problems everywhere
- Apple tests fundamentals hard — they care about clean code and edge cases
- Microsoft wants balanced skills — arrays, strings, and system thinking
This isn't speculation. It's data from 10,385 real interview questions asked across 259 companies, with timestamps showing what was asked in the last 6 months.
In this guide, you'll learn:
- What each top company actually asks (with data)
- Pattern preferences by company
- Difficulty mix expectations
- How to build a targeted 30–60 day prep plan
- How DSAPrep.dev tracks company-wise progress
The Data: What We Analyzed
We compiled 10,385 verified LeetCode questions tagged by company and timeframe:
| Company Category |
Companies |
Total Questions |
| FAANG/MAANG |
Amazon, Google, Meta, Apple, Microsoft |
2,322 |
| Banks & Quant |
Goldman Sachs, Bloomberg, Citadel, DE Shaw, JPM |
1,349 |
| Unicorn Startups |
DoorDash, Airbnb, ByteDance, Databricks |
581 |
| Top Tech |
LinkedIn, Adobe, Atlassian, IBM |
980 |
Timeframe breakdown:
- Last 3 months: 2,847 questions
- Last 6 months: 1,923 questions
- 6+ months ago: 5,615 questions
💡 Key Insight
Companies recycle core questions but the frequency and emphasis changes. Knowing what's hot right now gives you a measurable edge over candidates preparing blindly.
Company-Wise Breakdown: The Big 5 (FAANG/MAANG)
1. Amazon (500 Questions Analyzed)
Pattern DNA:
Top 5 Patterns:
1. Trees & Binary Search Trees (18%)
2. Arrays & Strings (16%)
3. Design / OOP (14%)
4. Graphs (12%)
5. Linked Lists (11%)
Difficulty Mix:
- Easy: 28%
- Medium: 58%
- Hard: 14%
What This Means:
Amazon loves practical problems. Expect:
- LRU Cache, design parking lot (they test real system thinking)
- Tree traversals (validate BST, lowest common ancestor)
- String manipulation with edge cases
🎯 Pro Tip
Amazon interviews mention Leadership Principles even in coding rounds. If asked "Tell me about a time...", they're testing Ownership, Bias for Action, and Customer Obsession — not just your code.
→ See Amazon Top 20 Questions
2. Google (500 Questions Analyzed)
Pattern DNA:
Top 5 Patterns:
1. Graphs & Advanced Search (22%)
2. Dynamic Programming (18%)
3. Arrays (15%)
4. Math & Logic (12%)
5. Backtracking (10%)
Difficulty Mix:
- Easy: 18%
- Medium: 52%
- Hard: 30%
What This Means:
Google wants you to optimize and scale. Expect:
- Graph problems (shortest path, connected components)
- DP with optimization (not just solve — improve from O(n²) to O(n))
- Follow-ups: "Can you do better?" "What if n = 10^9?"
🎯 Pro Tip
Google cares about complexity analysis more than perfect syntax. Be ready to discuss time/space trade-offs fluently — interviewers will push you on this every round.
→ See Google Top 20 Questions
Pattern DNA:
Top 5 Patterns:
1. Graphs (20%)
2. Hash Maps & Sets (17%)
3. Trees (15%)
4. Arrays & Strings (14%)
5. Backtracking (11%)
Difficulty Mix:
- Easy: 22%
- Medium: 60%
- Hard: 18%
What This Means:
Meta = social graphs. Everything connects. Expect:
- Friend recommendations (graph traversal)
- Newsfeed ranking (heaps, sorting, design)
- String manipulation (content moderation, search)
🎯 Pro Tip
Meta loves iterative problem-solving. They give you a basic problem, then progressively add constraints. Stay calm, adapt cleanly, and narrate your thinking out loud.
→ See Meta Top 20 Questions
4. Apple (375 Questions Analyzed)
Pattern DNA:
Top 5 Patterns:
1. Arrays & Strings (24%)
2. Trees (16%)
3. Linked Lists (14%)
4. Hash Maps (12%)
5. Sorting & Searching (11%)
Difficulty Mix:
- Easy: 32%
- Medium: 56%
- Hard: 12%
What This Means:
Apple tests fundamentals deeply. Expect:
- Classic problems done perfectly (reverse linked list, validate parentheses)
- Edge cases matter heavily (null pointers, empty arrays, integer overflow)
- Clean code — they care about readability and structure
🎯 Pro Tip
Apple interviewers ask about trade-offs even on easy problems. "Why hash map over sorting?" Know the reasoning behind every choice you make — not just the solution.
→ See Apple Top 20 Questions
5. Microsoft (447 Questions Analyzed)
Pattern DNA:
Top 5 Patterns:
1. Arrays & Strings (20%)
2. Dynamic Programming (16%)
3. Trees (14%)
4. Graphs (12%)
5. Linked Lists (11%)
Difficulty Mix:
- Easy: 26%
- Medium: 58%
- Hard: 16%
What This Means:
Microsoft is balanced and practical. Expect:
- DP problems (often with real-world context)
- System design awareness even in coding rounds
- String processing (Office, Azure scenarios)
🎯 Pro Tip
Microsoft loves real-world examples. Relate your solution to actual systems — caching, search, file systems. It shows product thinking, which they value as much as raw coding skill.
→ See Microsoft Top 20 Questions
Banks & Quant Firms: The Finance Tech Edge
Goldman Sachs (325 Questions)
Pattern Focus
- Arrays & Math (20%)
- Dynamic Programming (18%)
- Graphs (15%)
What's Different:
- Heavy math and logic problems (probability, combinations)
- More optimization questions (portfolio problems as DP)
- Behavioral + technical combo — they test culture fit hard
→ See Goldman Sachs Top 20 Questions
Bloomberg (433 Questions)
Pattern Focus
- Arrays & Strings (22%)
- Hash Maps (18%)
- Design (16%)
What's Different:
- Real-time data processing scenarios
- More design questions (terminal features, data feeds)
- Concurrency awareness even in DSA rounds
→ See Bloomberg Top 20 Questions
Citadel & DE Shaw (215 & 211 Questions)
Pattern Focus
- Math & Algorithms (24%)
- Dynamic Programming (20%)
- Graphs (14%)
What's Different:
- Competitive programming style — optimize heavily
- Probability and statistics embedded in problems
- Expect multiple rounds focused on algorithms only
→ See Citadel Top 20 Questions | → See DE Shaw Top 20 Questions
Unicorn Startups: Speed & Practicality
DoorDash (234 Questions)
Pattern Focus
- Graphs & Geospatial (26%)
- Arrays & Strings (18%)
- Design (16%)
Real-World Slant:
- Route optimization (Dijkstra, A*)
- Matching algorithms (drivers to orders)
- Real-time systems design
→ See DoorDash Top 20 Questions
Airbnb (132 Questions)
Pattern Focus
- Design & OOP (24%)
- Trees (18%)
- Arrays (16%)
Real-World Slant:
- Calendar/booking systems (interval problems)
- Search ranking (sorting, heaps)
- Recommendation engines (graphs, ML-adjacent)
→ See Airbnb Top 20 Questions
The 20 Most Asked LeetCode Questions Globally
Beyond company-specific patterns, some problems appear so frequently across all companies that mastering them is non-negotiable. These 20 problems were identified from our full dataset of 10,385 verified interview questions across 259 companies — ranked purely by total appearance count.
💡 Why This List Matters
These aren't just popular on LeetCode — they are the problems that keep showing up in actual interviews at Google, Amazon, Meta, Goldman Sachs, Citadel, and 250+ more companies. If you're short on time, master these 20 first.
What the Difficulty Mix Tells You
Of the top 20 globally asked problems, 7 are Easy, 11 are Medium, and only 2 are Hard. This confirms a key insight: companies don't gate-keep with Hard problems as often as people fear. They use well-known Medium problems to test how you think — your approach, your communication, and your ability to handle edge cases cleanly.
⚠️ Don't Skip the Easy Problems
Seven of the top 20 most asked problems are rated Easy — but companies use them to test code quality, edge case handling, and communication, not just correctness. A sloppy solution to Two Sum will cost you an offer just as much as a wrong solution to a Hard problem.
Pattern Breakdown of the Top 20
Two Pointers / Greedy 5 problems (#2, #7, #10, #11, #14, #16)
Dynamic Programming 3 problems (#10, #13, #17)
Binary Search 3 problems (#5, #12, #20)
Hash Map / Sets 3 problems (#1, #6, #9)
Matrix / Simulation 2 problems (#18, #19)
Stack 1 problem (#3)
Linked List 1 problem (#4)
Graphs / Union Find 1 problem (#8)
Strings 1 problem (#15)
✅ Takeaway: 4 Patterns Cover 75% of the Top 20
If you deeply master Two Pointers, Dynamic Programming, Binary Search, and Hash Maps, you can confidently attempt 15 of the top 20 most asked problems globally. These four patterns are your highest-ROI investment for any company interview.
How to Use This Data: Your 30–60 Day Company-Specific Plan
Step 1: Pick Your Target Companies (Max 3)
Don't try to prepare for 15 companies. Pick 1–3 targets based on where you're actually interviewing.
Example Split
- Primary: Google — focus 60% of time
- Secondary: Amazon — 30%
- Backup: Microsoft — 10%
Step 2: Analyze Pattern Overlap
Use DSAPrep.dev to filter by multiple companies and see pattern intersections. Google + Amazon overlap heavily on Trees, Graphs, and Dynamic Programming. Master those = prepare for both simultaneously.
Step 3: Solve Problems in This Order
Week 1–2 Core Patterns (40 problems)
- Solve top 10 problems per pattern
- Tag them in DSAPrep.dev with company names
- Schedule spaced repetition reviews
Week 3–4 Company-Specific Mix (40 problems)
- Solve actual company-tagged problems on LeetCode
- Mix difficulties: 30% easy, 60% medium, 10% hard
- Focus on last 6 months questions first
Week 5–6 Mock Interviews (20 problems)
- 2–3 mocks per week using company-specific problems
- Time yourself strictly (45 mins per problem)
- Record yourself explaining solutions
Step 4: Track Pattern Coverage
Use the DSAPrep.dev dashboard to see:
- ✅ Which patterns you've covered per company
- ⚠️ Which patterns you're weak on
- 📅 When to review company-specific problems
Common Mistakes in Company-Wise Prep
⚠️ Mistake 1: Only Solving Company-Tagged Problems
Why it fails: LeetCode company tags are often outdated or incomplete.
Better approach: Learn the
patterns the company favors, then solve any problem in that pattern. Pattern recognition > memorizing questions.
⚠️ Mistake 2: Ignoring Timeframes
Why it fails: A question asked 3 years ago may not reflect current interview trends.
Better approach: Prioritize
last 6 months data. Use older questions only for pattern practice.
⚠️ Mistake 3: Skipping Behavioral Prep
Why it fails: At Amazon, Apple, and banks, behavioral fit matters as much as coding.
Better approach: Prepare STAR stories for each company's values — Amazon LP, Apple's attention to detail, Goldman's analytical rigor.
⚠️ Mistake 4: Not Using Spaced Repetition
Why it fails: You solve 50 Amazon questions but forget 40 of them by interview day.
Better approach: Use DSAPrep.dev to schedule reviews. Solving once ≠ learning.
How DSAPrep.dev Helps: Company-Wise Tracking
1. Company Filters
- Filter problems by Amazon, Google, Meta, and more
- See which patterns each company emphasizes
- Track coverage % per company
2. Smart Review Scheduling
- Company-tagged problems auto-scheduled for review
- Prioritize high-frequency patterns per company
- "Due today" queue grouped by company
3. Pattern Heatmaps
- Visual chart: "Google asks 22% graph problems — you've done 8%"
- Identify gaps instantly
- Suggested problems to fill those gaps
4. Progress Dashboards
- "Amazon ready: 65%" — based on pattern coverage + review completion
- "Meta weak areas: Backtracking, Graphs"
- Weekly goals auto-generated based on your target interview date
Your Next Steps
Today
- Pick your top 3 target companies
- Sign up to DSAPrep.dev
- Add 5 problems per company to start tracking
This Week
- Analyze pattern overlap between your targets
- Solve 10–15 problems in high-overlap patterns
- Tag all problems with company names in DSAPrep
This Month
- Complete 60–80 company-specific problems
- Run 2–3 company-themed mock interviews
- Review all problems at least twice using spaced repetition
Conclusion: Stop Random Grinding, Start Strategic Prep
The difference between candidates who pass and those who don't isn't intelligence or hours spent. It's strategic focus.
Company-wise prep means:
- ✅ You solve problems that actually get asked
- ✅ You recognize patterns faster in interviews
- ✅ You use limited time efficiently
- ✅ You remember what you learn (with spaced repetition)