What is INP by TTFB
INP by TTFB shows how Interaction to Next Paint (INP) a measure of responsiveness — varies depending on Time to First Byte (TTFB), which reflects backend latency.
This lens connects server response times with user interaction smoothness. It helps answer an important question: Does backend slowness ripple into how responsive the site feels once it's visible?
Healthy INP by TTFB sample
Should you worry
In a healthy INP by TTFB view:
Fast TTFB sessions pair with fast INP.
The server responds quickly, and the frontend is responsive soon after.
Users experience minimal delay from load to interaction.
This indicates that your backend and frontend are working in sync to deliver a smooth, responsive experience.
Unhealthy INP by TTFB sample
Here’s where things go wrong:
Slow TTFB, slow INP: Classic full-stack drag. The server delays the HTML, and when it finally arrives, the frontend is still bogged down — maybe from JavaScript hydration, render-blocking assets, or poor thread scheduling.
Fast TTFB, slow INP: Backend is snappy, but the frontend can’t keep up. This often points to client-side issues — unoptimized JavaScript, unnecessary render work, or interaction handlers being blocked.
Slow TTFB, fast INP: Rare, but could indicate that the site is lightweight once it arrives. Still, the slow start risks user drop-off before they even reach interaction.
This lens is crucial for identifying whether slow interaction stems from the server side, the client side, or both.
Resolving unhealthy INP by TTFB
Go-to action plan to resolve an unhealthy INP by TTFB:
Ask Uxi to analyze your INP by TTFB values and suggest improvements.
Use Filters to isolate slow INP sessions tied to elevated TTFB. Identify recurring layout patterns or regions with latency spikes.
Simulate INP of the suspected lens to see if fixing it will resolve the INP by TTFB. If yes, this is where the resolution focus should be.
Use an automated INP optimization tool like INProve to improve your INP by TTFB values.
Once you’ve improved INP, set an alert to be the first to know if it starts worsening again.
Try it yourself
Discover how your website performs with real user data.