21 first-hand experiences · See open roles at Netflix
Netflix runs a senior-biased, team-specific loop: you are matched to a particular team before you interview rather than hired into a general pool, and a recruiter screen is followed by a hiring-manager conversation, a technical screen that is often a take-home instead of a live coding round, then an on-site round of engineer/manager interviews and a second round with directors and cross-team partners. The coding questions candidates describe are practical rather than puzzle-style (implement a TTL cache, a jq-like nested-map lookup, a transactional key-value store, an OOD feature class), and the take-homes are graded hard on judgment calls the brief never states. The distinctive part is the culture screening: the culture memo is sent to candidates ahead of the first call, opinions on it are actively probed, and multiple technically strong candidates report being rejected purely on the culture/behavioral round with no feedback.
These 21 writeups cover software engineering, design, engineering management roles, US (Bay Area/Los Gatos) heavy. 20 of them were posted under a pseudonym (Reddit, GeeksforGeeks or forum handles), so we could not verify the authors’ identities.
Most frequently reported · Behavioral(10), Data structures & algorithms(10), Past projects(8), Arrays & strings(4), Culture memo(4)
Netflix's own engineering blog (Karen Casella, Director of Engineering, Feb 2022) lays out the backend loop stage by stage: recruiter phone screen to check high-level fit and steer you toward a specific open role, then a hiring-manager phone screen on technical background and how you work, then a technical screen where many teams let you choose between a take-home exercise and a one-hour discussion with a team engineer. On-site Round 1 is four or five 45-minute sessions with two or three engineers plus the hiring manager and recruiter, using design and coding problems drawn from the team's real work; Round 2 is two or three more 45-minute sessions with an engineering director, a partner engineer or manager and another leader, focused on cross-team partnership and non-technical skills. The post explicitly advises against puzzle-style coding practice and says some orgs use a 'centralized hiring' variant with pivot points that can move you to a better-matched team instead of rejecting you. The separate culture memo at jobs.netflix.com/culture is the document candidates are told to read: dream team / keeper test, people over process, and the values (selflessness, judgment, candor, creativity, courage, inclusion, curiosity, resilience) that the behavioral rounds score against.
Read Netflix’s own hiring page →459 open roles in our index · advertised pay floors run $200k-$388k (median $300k, from 261 postings that state pay).
Behavioral·Low-level design·OOP·Data structures & algorithms·System design·Concurrency·Scalability·Past projects·Resume deep-dive·Domain knowledge·API design·Testing
Not selected1 round
The recruiter sent the culture memo to read before the very first call and then asked what the candidate thought of it. A half-joking answer that the document read as culty ended the process at that stage.
Covered · Behavioral, Culture memo
Shared by carom on hn · 2024-06-27
Not selected3 rounds
The candidate reports doing well across the technical portion of the loop, which was not what ended the process.
Covered · Data structures & algorithms, Past projects
A behavioral round with the hiring manager scored against the culture memo. The candidate had read the memo, got no feedback on what went wrong, and was passed on.
Covered · Behavioral, Culture memo, STAR stories
Recruiting re-slotted the candidate with a different team, which meant repeating the culture round with a new hiring manager. That one also ended in a rejection.
Covered · Behavioral, Culture memo
Shared by silenced_trope on hn · 2024-07-17
Not selected3 rounds
Unlike the other companies in the candidate's search, Netflix assigned him to a specific team before any interviews happened; the recruiter picked the team and the candidate was not shown alternatives.
Covered · Domain knowledge
The technical portion went well by the interviewers' own account.
Covered · Data structures & algorithms, Past projects
The rejection reason given was that the panel did not sense enough enthusiasm for the specific product area he had been earmarked for, despite strong technical performance.
Covered · Domain knowledge, Behavioral
Shared by bsimpson on hn · 2024-06-27
Outcome not shared1 round
The candidate was asked to open a real consumer interface of their choosing, invent a user persona for it, and argue on the spot about how well the interface served that persona's goals. A design lead pushed back on the critique by pointing to A/B testing behind the design decisions.
Covered · Product sense, User research, Case study
Shared by shadowfoxx on hn · 2023-12-01
Not selected1 round
A 24-hour take-home that the brief said should take two or three hours and warned against being clever. The candidate spotted that an interval tree was the only structure that would hit the performance target, judged it too clever, shipped a naive version with a note explaining the tradeoff, and was told the interval tree was in fact the expected answer.
Covered · Data structures & algorithms, Trees, Testing
Shared by kirbypineapple on hn · 2018-02-12
Not selected1 round
Build several movie-poster carousels without using a web framework, effectively rolling a small rendering layer in vanilla JS. The candidate spent around seven hours on a loosely time-boxed brief and was marked down for using the DOM as the source of truth instead of implementing a virtual DOM.
Covered · JavaScript, Front-end, Low-level design
Shared by suzzer99 on hn · 2022-06-21
Not selected1 round
A project-style exercise where the candidate reached for a state-machine library. The stated rejection reason was that the submission did not demonstrate enough command of plain JavaScript.
Covered · JavaScript, Front-end, Low-level design
Shared by inerte on hn · 2026-03-04
Not selected2 rounds
The loop opened with a take-home rather than a live algorithm screen, which the candidate (who has severe interview anxiety) found far closer to real work.
Covered · Past projects, Debugging
Every subsequent technical conversation was a high-level walkthrough and code review of the submitted take-home rather than fresh puzzle questions.
Covered · Past projects, API design, Testing
Shared by some-guy on hn · 2020-07-15
Not selected1 round
The hiring manager's feedback was that the submission targeted an older Java version, which was read as a signal that the candidate does not keep current with the ecosystem.
Covered · Java, Past projects
Shared by dangwhy on hn · 2023-04-08
Received an offer3 rounds
First contact came directly from the hiring manager rather than a recruiter, and an on-site was booked three days later.
Covered · Resume deep-dive, Past projects
The first on-site phase, one of two, with the second booked three days after it.
Covered · System design, Past projects, Behavioral
The second on-site phase, after which the candidate was given a verbal offer on the way out of the building, above the number he had asked for. A separate comment from the same author puts the total at eight interviews.
Covered · System design, Behavioral
Shared by jedberg on hn · 2020-02-24
Received an offer1 round
The candidate was interviewed by the two managers whose teams he could join, and had repeated recruiting conversations about exactly what the job would be and how his existing industry experience would be used, rather than generic academic questions.
Covered · Past projects, Domain knowledge, Resume deep-dive
Shared by brendangregg on hn · 2017-02-21
Outcome not shared1 round
No coding was asked of the manager candidate. The conversations covered partnering across teams, coaching and working with engineers, setting technical direction, and handling customer and vendor relationships.
Covered · Behavioral, Conflict handling, System design, Past projects
Shared by dandigangi on hn · 2022-06-09
Not selected1 round
A conventional Bay Area on-site of whiteboard data-structure and algorithm questions. The interviewer told the candidate to his face at the end of the day that he had not passed, rather than leaving him to wait for a recruiter email.
Covered · Data structures & algorithms, Arrays & strings
Shared by jcadam on hn · 2020-09-28
Outcome not shared3 rounds
An on-site slot with the head of talent that was largely a monologue about the share price and about past firings, which read to the candidate as a demonstration of how central the talent function is to the company.
Covered · Behavioral, Domain knowledge
A prospective teammate gave a fizzbuzz-style problem and then asked for it to be rewritten repeatedly under added constraints. Missing one Python idiom produced feedback that the candidate lacked depth in computer science.
Covered · Arrays & strings, Python, Debugging
The panel generally liked the candidate and proposed moving him to a data warehouse team, a role quite different from the one he had been recruited for.
Covered · Domain knowledge
Shared by RayVR on hn · 2015-01-20
Not selected1 round
The candidate interviewed twice, years apart, and was turned down both times on culture grounds rather than technical ones: once with a vague reference to not fitting the culture deck, once with the explicit 'brilliant jerk' label, despite an internal referral from a former colleague.
Covered · Behavioral, Culture memo, Conflict handling
Shared by iamleppert on hn · 2016-06-28
Not selected4 rounds
A CodeSignal assessment on which the candidate solved roughly two and a half of the four problems and still advanced.
Covered · Data structures & algorithms, Arrays & strings
A conversation with a recruiter centred on the culture memo plus standard behavioral questions.
Covered · Behavioral, Culture memo
A single list-manipulation problem (reordering nodes so odd-indexed ones precede even-indexed ones), solved quickly with edge cases checked, followed by an informal chat about the team.
Covered · Linked lists, Data structures & algorithms
A relaxed conversation about past experience and interests. The candidate felt the rapport was weaker than in the recruiter round and received a non-automated rejection about five days later.
Covered · Past projects, Behavioral
Shared by LeetCode Discuss (author handle not shown) on leetcode · 2021-03-04
Not selected1 round
A CodeSignal test of four problems in 60 minutes, mixing easy array manipulation with a matrix diagonal-pattern search and a harder subarray counting problem. The candidate fully solved two and partially passed the rest, and was rejected.
Covered · Arrays & strings, Dynamic programming, Data structures & algorithms
Shared by fuck-the-luck on leetcode · 2022-01-20
Not selected3 rounds
An online coding assessment that the candidate passed.
Covered · Data structures & algorithms
A technical phone screen that also went well.
Covered · Data structures & algorithms
A three-interview final round split across two dates. After the first two sessions the candidate was told the headcount had been filled and was cut before the third, then got no reply to follow-up. The whole process had run about three months.
Covered · System design, Behavioral
Shared by LeetCode Discuss (author handle not shown) on leetcode · 2022-06-04
Outcome not shared2 rounds
A medium-difficulty exercise framed as designing a small feature: define a class and its methods from a set of requirements, using ordinary data structures rather than an algorithmic trick.
Covered · Low-level design, OOP, Data structures & algorithms
The recruiter described the on-site as two coding rounds, one problem-solving round and two behavioral rounds.
Covered · Data structures & algorithms, Behavioral, Problem solving
Shared by LeetCode Discuss (author handle not shown) on leetcode · 2024-09-17
Not selected1 round
Build a key-value store supporting transactions, required to be written in JavaScript. The candidate solved it with a map plus a queue but felt the solution was over-engineered, and did not advance.
Covered · Low-level design, Data structures & algorithms, JavaScript
Shared by LeetCode Discuss (author handle not shown) on leetcode · 2022-10-14
Outcome not shared1 round
A generally positive loop where several sessions were run by two interviewers at once, which the candidate found kept things civil and let more ground be covered. One coding session was an outlier: the interviewer could not articulate the task and changed the requirements repeatedly.
Covered · Concurrency, Scalability, Data structures & algorithms
Shared by Strunevskiy on leetcode · 2024-09-20
These guides summarize public, first-hand interview experiences shared by candidates. They describe the shape of each loop and the topics that came up, not a leaked question bank, and every experience links back to its original source. Processes change often; treat this as directional prep, not a script.