Roadmaps / Frontend / Interview topics
Reference
Interview topics
Use these for frontend and full-stack roles. Each follow-up is the question a good interviewer asks next. The Staff line is what lifts the answer.
Phase 1
| Question | Follow-up | A Staff answer adds |
|---|---|---|
| What happens between typing a URL and first paint? | Where can it be slow? | The critical path with numbers and how to read it in a trace |
| How does the event loop order work? | What blocks it? | Microtasks against tasks and a long-task example from a trace |
| What are the Core Web Vitals? | Why the 75th percentile? | Field against lab data and what you would do first on a failing page |
| Why does layout shift happen? | How do you prevent it? | Reserved space, font loading choices and a measured before and after |
System design prompt (20 minutes): a product listing page for a slow mobile network. Outline: targets and numbers (3 min), critical path (5), rendering and caching (5), failure cases (4), measurement (3).
Phase 2
| Question | Follow-up | A Staff answer adds |
|---|---|---|
| When do you use an effect? | What do you use instead? | A list of the cases that need one, and derived values for the rest |
| Where does state live? | How do you avoid duplicates? | One owner per piece of state and a rule for server against client state |
| How do you make a custom widget accessible? | When do you not use ARIA? | Native elements first, a keyboard map and a screen reader test |
| How do you adopt strict types in an old codebase? | How do you stop regression? | A staged plan, a ratchet in CI and a count of any over time |
System design prompt (20 minutes): a design system for five product teams. Outline: goals (3 min), component API and accessibility (5), versioning and releases (4), adoption and support (4), measuring outcomes (4).
Phase 3
| Question | Follow-up | A Staff answer adds |
|---|---|---|
| Server, static or client rendering? | How do you decide per route? | A table with measured reasons and a caching rule |
| How do you improve INP? | What if the long task is third party? | Trace evidence, scheduling work and a budget for third-party scripts |
| How do you keep bundle size down over time? | Who owns the budget? | A gate in CI, an owner and a review for each new dependency |
| How do you find a memory leak? | How do you stop it returning? | Snapshots, a test that repeats an action and a monitor |
System design prompt (20 minutes): speed up a slow checkout page. Outline: baseline and targets (3 min), trace and diagnosis (5), fixes in order of gain (5), guardrails (4), rollout (3).
Phase 4
| Question | Follow-up | A Staff answer adds |
|---|---|---|
| How do you prevent cross-site scripting? | What does a CSP add? | Output encoding by context, a strict policy and a test for a payload |
| How do you make end-to-end tests reliable? | What about flaky ones? | Waiting on state, isolated data and a quarantine rule with an owner |
| How do you roll out a risky change? | How do you roll back? | A flag, a staged rollout, a metric and a tested off switch |
| How do you keep up with a fast ecosystem? | When do you upgrade? | A schedule, a trial branch, a dependency budget and a support window |
System design prompt (20 minutes): a frontend platform for ten teams. Outline: goals (3 min), shared tooling and templates (5), quality gates (5), upgrades and support (4), measuring outcomes (3).