Web Vitals Checklist

Framework-agnostic performance criteria. Focuses on what to verify, not how to measure.

Core Web Vitals Targets

• •Pass: Google official threshold. No SEO penalty.
• •Excellent: Top-tier site performance. Unofficial target.

1. Loading Performance (LCP)

Critical Path

• • No render-blocking resources in document head
• • Critical resources load before non-critical
• • Document response starts within TTFB target

LCP Element

• • LCP element identified and loads early
• • LCP element not lazy-loaded
• • LCP image prioritized over other resources
• • No client-side rendering delays for LCP content

Resource Loading

• • Critical third-party origins connected early
• • Above-fold images load without delay
• • No unnecessary redirects in critical path

2. Interactivity (INP)

Main Thread

• • No long tasks (> 50ms) during interaction
• • Heavy computation offloaded from main thread
• • Event handlers complete within 100ms

Input Response

• • Visual feedback appears immediately on interaction
• • Input events not blocked by other scripts
• • Frequent events (scroll, input) properly throttled

Script Execution

• • Third-party scripts don't block user interaction
• • No synchronous storage access in interaction handlers
• • DOM updates batched efficiently

3. Visual Stability (CLS)

Layout Reservation

• • Images have dimensions before load
• • Videos and iframes have reserved space
• • Ads and embeds don't shift content
• • Dynamic content has minimum space reserved

Font Stability

• • Text doesn't reflow after font load
• • Fallback font size matches web font
• • No invisible text during font load (unless intended)

Content Injection

• • No content inserted above existing content
• • Animations don't trigger layout shifts
• • Skeleton/placeholder matches final content size

4. Network & Caching

Cache Effectiveness

• • Static assets cached long-term
• • Cache hit rate > 80% for repeat visits
• • API responses cached appropriately
• • No unnecessary network requests on repeat visits

Compression

• • Text resources compressed (Brotli preferred)
• • Compression ratio reasonable for content type

Protocol

• • HTTP/2 or HTTP/3 in use
• • Connection reuse effective
• • No excessive parallel connections to same origin

CDN

• • Static assets served from edge locations
• • Cache headers set correctly at edge
• • Geographic latency acceptable for target users

5. Images

Format Efficiency

• • Modern formats used where supported (WebP, AVIF)
• • Format appropriate for content type
• • No unnecessarily high quality settings

Sizing

• • Images not larger than display size
• • Responsive images serve appropriate sizes
• • No oversized images on mobile

Loading Behavior

• • Below-fold images lazy-loaded
• • Above-fold images load immediately
• • No layout shift when images load

6. JavaScript

Bundle Size

• • Total JS size within budget
• • No duplicate dependencies in bundle
• • Dead code eliminated

Loading Behavior

• • Critical JS loads first
• • Non-critical JS deferred
• • Third-party scripts don't block page load

Code Splitting

• • Routes load only required code
• • Large features split into separate chunks
• • Initial bundle minimal

Dependencies

• • No unused dependencies in bundle
• • Heavy libraries loaded only when needed
• • No full library imports when subset sufficient

7. CSS

Delivery

• • Critical CSS available immediately
• • Non-critical CSS doesn't block render
• • Total CSS size within budget

Efficiency

• • No unused CSS in critical path
• • Selector complexity reasonable
• • No excessive specificity wars

Animation Performance

• • Animations don't trigger layout recalculation
• • Animations run at 60fps
• • Off-screen content rendering optimized

8. Fonts

Loading Impact

• • Fonts don't block text rendering excessively
• • Font files load within LCP budget
• • No visible font swap flash (or acceptable)

Efficiency

• • Font files reasonably sized
• • Only necessary font weights loaded
• • Font family count minimal

9. Third-Party Scripts

Audit

• • All third-party scripts justified and necessary
• • Third-party impact measured and acceptable
• • Unused third-party scripts removed

Performance Impact

• • Third-party scripts don't block page load
• • Third-party scripts don't degrade INP
• • Third-party content doesn't cause CLS

Resilience

• • Page functional if third-party fails
• • Third-party timeouts don't block critical path

10. Server Performance

Response Time

• • Server response within TTFB target
• • Database queries not blocking response
• • Server-side caching effective

Infrastructure

• • Server located near target users (or edge deployed)
• • Server resources adequate for load
• • No bottlenecks under normal traffic

Monitoring

• • Real User Monitoring (RUM) active
• • Core Web Vitals tracked in production
• • Performance regression alerts configured

Performance Budget