Streaming Metrics
You'll learn how to measure performance when your server sends HTML in progressive chunks rather than waiting for the complete response.
What Are Streaming Metrics?
When using streaming SSR, your server doesn't wait until the entire page is built before sending it. Instead, it sends HTML in chunks as they become ready—like streaming a video rather than downloading it completely first. This creates unique performance characteristics that traditional metrics don't fully capture.
Time to First Chunk (TTFC)
This measures how long it takes from the initial request until the first chunk of HTML arrives at the browser. Think of it like the moment a streaming movie starts playing—you haven't seen the whole thing, but something useful is happening immediately.
TTFC is crucial because it determines when the browser can start parsing HTML and making additional resource requests (CSS, JavaScript, fonts). A low TTFC means users see something faster, even if the full page isn't ready.
Progressive Rendering Metrics
With streaming, content appears in stages. You need to track:
- Time to first visible content — when users see meaningful UI (not just a loading shell)
- Time to streaming completion — when the last chunk arrives
- Chunk delivery intervals — how smoothly chunks arrive (gaps create visual stuttering)
Unlike a single LCP measurement, streaming may have multiple contentful paints as each chunk renders. You might track "time to header," "time to main content," and "time to footer" separately.
Why This Matters
Streaming metrics help you understand if your progressive rendering strategy actually feels faster to users. A page that shows the navigation in 200ms, then main content at 400ms, then footer at 600ms can feel more responsive than one that shows everything at 500ms—even though the total time is longer.
Key Takeaway: Streaming SSR requires measuring progressive delivery with Time to First Chunk and staged content appearance metrics, not just traditional full-page load times.