Web Performance Roadmap: Start Here (2 hours / 1 day / 1 week)
A guided entry point into the Web Performance roadmap with curated paths (2 hours / 1 day / 1 week) + exactly what to measure and why it matters in interviews.
Frontend Interview Team
March 01, 2026
If you’re preparing for frontend interviews (or you’re owning performance for a real production app), you don’t need 50 random tips.
You need a sequence:
- Know what to measure (and what “good” looks like)
- Find the bottleneck (don’t guess)
- Fix the critical path (LCP) and interaction path (INP)
- Re‑measure and keep guardrails
This page gives you the best path through the Web Performance roadmap.
Quick definitions (so you don’t get lost)
- LCP (Largest Contentful Paint): when the main content looks ready.
- INP (Interaction to Next Paint): how responsive the page feels when users click/type.
- CLS (Cumulative Layout Shift): how stable the layout is while loading.
A senior way to talk about performance is: “What’s the critical path for LCP?” and “What’s the hot path for interactions?”
Choose your path
Path A — 2 hours (highest ROI)
If you only have 2 hours, do these in order:
- #01 Core Web Vitals: The Mental Model (LCP, INP, CLS)
- #02 Performance Debugging Playbook (Chrome DevTools + Real Signals)
- #09 Next.js Performance (RSC, streaming, caching)
- #05 JavaScript Cost (hydration, long tasks, shipping less JS)
- #12 Web Performance Interview Pack (top 25 Q&A)
You’ll be able to answer most performance questions with confidence, and more importantly: you’ll have a debugging loop.
Path B — 1 day (solid interview prep)
Do Path A, plus:
- #03 Images for Performance (sizes, formats, priority, preload)
- #04 Fonts Without Regret (FOIT/FOUT, preload, CLS-safe loading)
- #06 Caching & CDNs (HTTP caching, validation, edge)
- #07 Rendering Pipeline (layout → paint → composite)
- #10 Measuring Performance (Lighthouse vs RUM, budgets)
Path C — 1 week (production-ready)
Do Path B, plus:
- #08 Third-Party Scripts (tag manager hygiene)
- #11 Performance Recipes (before/after case studies)
- #13 CSS Performance (containment, content-visibility)
- #14 Performance APIs (Long Tasks, Web Vitals, observers)
- #15 Offline & Service Workers (caching strategies)
What to measure (minimum setup)
If you’re interviewing, you should always be able to say:
- I measured it (not “I think it’s slow”)
- I located the bottleneck in DevTools
- I validated the fix (lab + field if possible)
Minimum metrics to track
- LCP and the LCP element (what actually became the LCP)
- INP (or at least long tasks + interaction delays)
- CLS
- TTFB (especially for SSR apps)
Minimum tooling
- Chrome DevTools: Performance + Network
- Lighthouse (lab) + a RUM signal (field) if possible
A senior debugging loop (copy/paste)
When someone says “this page is slow”, do this:
- Reproduce on a slower profile (CPU 4× / 6×, cache off on first run)
- Record a Performance trace while doing the real user action
- Find the first bottleneck:
- LCP blocked by TTFB, render-blocking, or late resource discovery
- INP blocked by long tasks, rerenders, or expensive handlers
- Fix in priority order:
- shorten the critical path
- ship less JS / reduce hydration
- reduce rerenders
- move work off the main thread
- Re‑measure and add a guardrail
Roadmap (1 → 15)
- #01: /blog/web-performance-core-web-vitals-mental-model
- #02: /blog/web-performance-debugging-playbook-chrome-devtools
- #03: /blog/web-performance-images-sizes-formats-priority-preload
- #04: /blog/web-performance-fonts-fast-without-cls
- #05: /blog/web-performance-javascript-cost-hydration-long-tasks
- #06: /blog/web-performance-http-caching-and-cdn
- #07: /blog/web-performance-rendering-pipeline-layout-paint-composite
- #08: /blog/web-performance-third-party-scripts-containment
- #09: /blog/web-performance-nextjs-rsc-streaming-caching
- #10: /blog/web-performance-lighthouse-vs-rum-and-budgets
- #11: /blog/web-performance-recipes-high-roi-before-after
- #12: /blog/web-performance-interview-pack-25-questions
- #13: /blog/web-performance-css-performance-containment-content-visibility
- #14: /blog/web-performance-apis-long-tasks-web-vitals
- #15: /blog/web-performance-offline-service-worker-caching
What’s next after this roadmap?
If you finish these lessons, the next best “performance multiplier” skills are:
- React rendering fundamentals (rerenders, memoization, concurrency)
- Next.js architecture (RSC boundaries, caching, streaming)
- Observability (RUM + tracing + performance budgets)