Time to First Byte

What this check looks for

For pages with measured performance data, this check grades TTFB — the time between the request being sent and the first byte of the response arriving — against a good threshold of 800ms and a poor threshold of 1800ms. At or under 800ms passes; between 800ms and 1800ms warns; beyond 1800ms fails. Pages without measured TTFB report "na".

Why it matters

TTFB is the floor under every other performance metric on the page: no rendering, no script execution, and no Largest Contentful Paint can begin until the server has sent back its first byte of HTML, so a slow TTFB pushes back everything downstream of it by the same amount. It's typically driven by server-side factors — slow database queries, missing caching, cold-start delays, or geographic distance between the visitor and the server — rather than anything in the page's own markup. Faster TTFB compounds directly into faster LCP, which is a scored Core Web Vital, and a slow-responding server is also more likely to time out or be deprioritized by AI crawlers working under their own fetch budgets.

How to fix it

Add server-side caching and reduce backend work per request:

# Cache-Control on server-rendered HTML that doesn't change per-request
Cache-Control: public, max-age=60, stale-while-revalidate=300

# Or serve through a CDN with edge caching close to visitors

Profile what the server actually does before responding — a slow database query or an uncached API call in the render path is the most common root cause of a high TTFB.