What Big Tech Interview Loops Actually Test

Big tech interview loops have more in common than the folklore suggests. Every one of them is four or five rounds, two or three of which are coding, one system design above a certain level, and one behavioural. The differences are real but they are differences of emphasis, and knowing which emphasis to expect is worth more than any list of questions.

The shape they share

Where the emphasis differs

These are tendencies drawn from published hiring material and widely reported candidate experience. They are useful for deciding what to over-prepare, not for predicting questions.

Company	Documented emphasis	Over-prepare
Google	Algorithmic depth, hiring committee review	Justifying complexity out loud
Amazon	Leadership Principles, published openly	Behavioural stories mapped to named principles
Meta	Speed, two problems in one round	Finishing cleanly under time pressure
Microsoft	Practical engineering, team fit	Talking about real systems you have built
Apple	Team-specific and domain deep	Depth in the specific area on the job posting
Netflix	Seniority and judgement, published culture memo	Decisions you owned end to end

The pattern across the table is that only one column is company-specific. The coding round is broadly interchangeable, which is why preparation that generalises is more efficient than preparation targeted at one logo, right up until the fortnight before the loop.

"A language that doesn't affect the way you think about programming is not worth knowing."
- Alan Perlis, Epigrams on Programming (1982)

The corollary for interviews: preparation that only changes what you can recall is not worth much. Preparation that changes how you approach an unfamiliar problem in front of a stranger transfers to every loop on the table above.

The round that eliminates most people

The technical phone screen removes more candidates than the entire onsite, and it is the round people prepare for least seriously because it feels like a formality. It is harder than the onsite in one specific way: there is no whiteboard, no body language, and a single moderate problem where any stumble is a large fraction of the evidence.

Do not skip the behavioural round

Engineers routinely allocate ninety-five percent of preparation to the technical rounds and then lose the offer on the one where they had no stories ready. At senior levels the behavioural round is not a formality at all; it is often where the level is decided, and the level is where the money is.

Six prepared stories, each with a real decision you owned and at least one number, cover almost every version of that round across every company in the table. It is perhaps three hours of work for the round with the highest ratio of outcome to effort in the whole loop.