The debate is endless on Reddit and Twitter. We break down what actually matters for product startups vs service companies vs FAANG — and how to split your time.
This question floods every engineering college WhatsApp group in the months leading up to placement season. Should I grind LeetCode or build projects? The honest answer: it completely depends on where you want to work. The mistake most students make is treating this as a universal question with a universal answer.
If your target is Google, Microsoft, Meta, or similar companies, DSA is non-negotiable. These companies compete on algorithmic efficiency at scale, and their interviews are specifically designed to test your ability to reason about time and space complexity under pressure.
You need depth, not breadth. Aim for 200–300 LeetCode problems with genuine understanding, not blind memorization. Focus specifically on trees and graphs, dynamic programming, and system design. The system design round becomes critical for SDE-2 and above, but even as a fresher you'll be asked basic scalability questions.
Projects still matter here — they're your ticket into the conversation. But for FAANG, a thin project with strong DSA skills beats a thick portfolio with weak fundamentals. The ratio roughly: 60% DSA, 40% projects and system design basics.
Early-stage startups (under 100 people, especially Series A and earlier) almost never run whiteboard DSA interviews. They don't have time, and frankly, they're testing a different thing: can you ship?
At a 20-person startup, your first week might involve debugging production issues, adding a feature to a live codebase, and deploying something. None of that requires you to implement a red-black tree from scratch. It does require you to read unfamiliar code, understand system constraints, and make reasonable decisions under ambiguity.
For startup roles, focus on projects that demonstrate shipping ability. A full-stack app you've deployed and maintained. A side project with real users. Contributions to open source. System design basics (databases, APIs, caching) matter more than algorithmic puzzles. Ratio: 20% DSA basics, 80% projects and fundamentals.
Service companies hire at massive volume and optimize for a different profile. Their assessments test aptitude, basic programming logic, and communication — not advanced algorithms or impressive projects.
The bar here is fundamentals: can you write a working loop, understand basic OOP, and explain what a database join does? The interviews are structured, the questions are predictable, and preparation is mostly about practicing their specific test formats.
Communication and attitude often weigh more heavily in service company HR rounds than technical depth. Ratio: 30% basic DSA and aptitude, 70% fundamentals, communication, and getting comfortable with structured interview formats.
Before you open LeetCode or start a new project, spend 20 minutes writing down a list of 10–15 companies you'd actually want to work at. Be honest. Then group them by type: FAANG/product giants, growth-stage startups, early-stage startups, service companies.
That grouping tells you your split. If 8 of your 15 companies are startups, spend 80% of your time building and shipping. If 10 are product giants, grind the DSA harder than anyone in your batch.
The worst thing you can do is try to do both equally. You end up with 150 mediocre LeetCode solves and a half-finished project — not competitive for either track.
"There's no universal answer, but there is a personal one — and it takes about 20 minutes to find. Figure out where you want to work, and let that dictate everything else."
Career tactics, technical deep-dives, and honest advice — weekly.
