Performance patterns

Performance Patterns

Frontend Performance

Lazy Loading

const HeavyComponent = lazy(() => import('./HeavyComponent'))

Load components only when needed. Use for routes, modals, below-fold content.

Memoization

const memoizedValue = useMemo(() => computeExpensiveValue(a, b), [a, b])
const memoizedFn = useCallback(() => doSomething(a, b), [a, b])

Cache expensive computations and function references.

Code Splitting

Split your bundle by route or feature. Users only download what they need.

Image Optimization

• Use WebP/AVIF formats
• Lazy load below-fold images
• Use responsive images (srcset)
• Compress images (tinyPNG, Sharp)

CSS Performance

• Use transform and opacity for animations (GPU-accelerated)
• Avoid layout triggers (width, height, margin, padding)
• Use will-change sparingly
• Minimize repaints with contain

Backend Performance

Caching

• Cache database queries (Redis, in-memory)
• Cache API responses (HTTP caching headers)
• Cache expensive computations (memoization)
• Cache rendered pages (SSR cache)

Database Optimization

• Add indexes for frequent queries
• Use connection pooling
• Batch inserts/updates
• Avoid N+1 queries
• Use EXPLAIN ANALYZE

Async Operations

• Don't block the event loop
• Use worker threads for CPU-intensive tasks
• Stream large responses
• Queue background jobs

Measurement

• Lighthouse — overall performance score
• Core Web Vitals — LCP, FID, CLS
• Chrome DevTools — Performance tab, Lighthouse tab
• webpack-bundle-analyzer — bundle size visualization
• ab / wrk — load testing

Common Anti-Patterns

• Loading everything upfront
• Not caching expensive operations
• Synchronous database calls in loops
• Large bundle sizes (no code splitting)
• Unoptimized images (no lazy loading, wrong format)
• Layout thrashing (reading layout then writing)