The Interview Loop
Before the first problem, understand the machine you’re inside. Big-tech hiring is a pipeline of independent gates, each filtering for a different thing, each scored by a different person who hasn’t seen your other rounds. Knowing what each gate wants lets you spend your energy where it counts.
The five gates
Recruiter screen (~30 min)
Non-technical. The recruiter confirms your background, explains the role and timeline, and sniffs out comp expectations and red flags. You almost can’t “win” here — but you can lose by being unprepared, evasive about comp, or flaky on scheduling. Be warm, concise, enthusiastic about this role, and have 2–3 thoughtful questions ready.
Technical phone screen (~45–60 min)
The real filter, and where most candidates are cut. One or two coding problems (usually a medium) in a shared editor — often no autocomplete and no “run” button, so your code has to compile in your head. The bar: a working, reasoned solution communicated clearly. Pass this and you reach the onsite.
Onsite / virtual onsite (4–6 rounds)
The main event, back-to-back ~45-minute rounds:
- 2 coding rounds — harder mediums / easy-hards, sometimes two problems per round.
- 1 system design — expected for mid-level and senior (and the whole bar for senior). See the system design round.
- 1–2 behavioral — “tell me about a time…” mapped to company values. See the behavioral round.
- Sometimes a domain round — OS, ML, frontend, depending on the role.
Each interviewer scores you independently. You reset and re-earn the room every 45 minutes.
Hiring committee / debrief (async)
Interviewers write detailed feedback and vote on a scale (Strong Hire → Hire → Lean Hire → Lean No-Hire → No-Hire → Strong No-Hire). At Google a committee (who never met you) reads the packets; elsewhere the hiring manager reconciles. One confident “No-Hire” on a core signal can sink an otherwise strong loop — consistency across rounds matters.
Offer & negotiation (days)
The recruiter calls with numbers. Compensation is almost always negotiable — a competing offer, market data, and a calm, specific counter routinely move base, equity, and sign-on. Never accept on the spot; get the full breakdown in writing and take time to think.
Every onsite round starts from zero. Your brilliant first round doesn’t carry over — the next interviewer usually hasn’t read it, by design, to keep evaluations independent. That’s good news after a rough round (the next one is a clean slate) and a warning after a great one (you still have to perform). Treat each 45 minutes as the whole interview.
The same loop, weighted differently
FAANG/MAANG (Meta, Apple, Amazon, Netflix, Google) and the broader giants — Microsoft, and the rest of “MONGOS”-tier big tech — run a structurally similar loop. What changes is the weighting. Prep to the emphasis of the company in front of you:
Amazon is the big outlier — behavioral is not a formality, it’s the spine. Every Amazon round, even coding ones, probes the 16 Leadership Principles, and a “Bar Raiser” interviewer (from outside the team) can veto an offer. If you’re interviewing at Amazon and you’ve only practiced LeetCode, you’ve prepared for half the loop. Walk in with 6–8 STAR stories pre-tagged to LPs like Ownership, Customer Obsession, Dive Deep, and Bias for Action. (More on the behavioral page.)
How to allocate your prep
A rough budget for the typical software role, given limited time:
| If the loop is… | Spend most on |
|---|---|
| New grad / junior | Coding + DSA (the first 27 days). System design is usually light or absent. |
| Mid-level (most common) | ~60% coding, ~25% system design, ~15% behavioral. |
| Senior (L5/E5+) | System design becomes make-or-break; behavioral (scope, influence, leadership) rises sharply; coding bar stays but is rarely the deciding signal. |
| Amazon (any level) | Add a heavy behavioral block — STAR stories tagged to Leadership Principles. |
Quick check
Next: The Coding Round — the minute-by-minute framework, an interactive phase stepper, and a real practice timer.