performance-score

What this datapoint measures

Page-load and rendering performance, typically measured against Core Web Vitals (Largest Contentful Paint, First Input Delay or Interaction to Next Paint, Cumulative Layout Shift) and similar performance metrics.

Performance matters for AI-mediated discovery for two reasons. First, AI crawlers have time budgets per page; pages that exceed the budget may be skipped, partially indexed, or cached in degraded form. Second, performance affects user experience post-discovery — when an AI cites a brand’s page and the user follows the link, slow pages produce poor user experience that may downstream affect the brand’s reputation in retrievable signals.

What high looks like

• LCP under 2.5 seconds on representative pages
• INP under 200 milliseconds
• CLS under 0.1
• Server response time reasonable
• Render-blocking resources minimized
• Image optimization in place (modern formats, responsive images, lazy loading where appropriate)
• Critical CSS inlined where the optimization makes sense

What low looks like

• LCP over 4 seconds
• INP over 500ms
• CLS over 0.25
• Multi-second server response times
• Multiple render-blocking JavaScript and CSS resources
• Unoptimized images at full resolution
• Pages requiring extensive JavaScript execution before content appears

What at floor looks like

A brand at floor on performance-score has pages that take many seconds to render meaningful content, produce significant layout shift during rendering, and have poor interaction responsiveness. AI crawlers may receive partial content (the early-rendering parts) while later content fails to appear within the crawl budget.

This pattern is common in brands using heavy WordPress themes with many plugins, in brands using client-side rendered frameworks without server-side rendering or static generation, and in brands hosting on undersized infrastructure. The remedy is engineering-significant: performance work involves trade-offs and often touches the rendering architecture.

What affects this datapoint

• Server response time
• Time to First Byte (TTFB)
• Render-blocking resource count and size
• Total page weight (HTML + CSS + JS + images + fonts)
• Image optimization
• Font loading strategy
• Client-side JavaScript execution time
• Cumulative Layout Shift from late-loading content
• Third-party script impact (analytics, ads, embedded content)

OMG actions that influence this datapoint

Action	Influence
O-4 Technical Infrastructure, Performance & International Foundation	Direct, primary. Performance optimization is a core deliverable of O-4.
M-7 Multimedia Content Optimization	Substantial. Image and video optimization directly affects performance.

Multilingual considerations

Performance is language-neutral in measurement but can have language-specific implications:

• CJK fonts are larger than Latin fonts; pages rendering Japanese, Korean, or Chinese content with custom web fonts have a heavier font payload than equivalent English pages
• Right-to-left language handling adds rendering complexity (not relevant for current calibrated five primary languages)
• Geographic distribution of audiences for non-English brands may produce different performance experiences (a Japanese site served from a US-only CDN performs differently than the same site served from a CDN with Asia-Pacific edges)

Common failure modes

• WordPress sites with too many plugins; cumulative JavaScript and CSS becomes performance-killing
• Page builders generating heavily nested DOM with extensive inline styles
• Hero images served at desktop resolution to mobile devices
• Custom fonts loaded with font-display: block causing flash of invisible text
• Third-party tracking scripts that block rendering
• CDN cache misses producing slow origin-server responses
• Single-page applications without server-side rendering or static generation; AI crawlers see the empty shell

Diagnostic interpretation

Performance-score at low with crawlability also low suggests the performance issues are causing crawl failures. Remediation is doubly urgent.

Performance-score at low with crawlability high indicates performance issues that affect users but not crawlers (because crawlers have sufficient time budget). Remediation is still appropriate but less urgent for AVO specifically; it is more about user experience post-citation.

Performance-score at high with other V1.2 datapoints at low indicates a brand with fast-loading pages but other infrastructure issues. The brand has invested in performance specifically (often via CDN and caching) without addressing other technical health concerns.

---