What Interaction to Next Paint measures

How it works

INP is one of the three Core Web Vitals and replaced the older First Input Delay in March 2024. For every interaction during a visit, the browser measures the time from the input to the next frame it draws. That time has three parts:

1. Input delay: the main thread is still busy with something else, usually a script.
2. Processing: the scripts that respond to the input run.
3. Presentation: the browser recalculates the layout and draws.

INP is roughly the slowest interaction of a visit; with very many interactions, single outliers are left out. Up to 200 ms is good, over 500 ms counts as poor, each for three out of four page views.

INP can only be measured with real visitors. Nobody clicks in the lab; there, Total Blocking Time (TBT) stands in for it: how long long tasks, meaning scripts running for more than 50 ms in one go, block the main thread while the page loads. bytereport rates it good up to 200 ms and poor over 600 ms.

What matters

• Less JavaScript. What doesn't run doesn't block, see JavaScript.
• Split long tasks. A script that computes for 300 ms in one go blocks every input during that time. Split into small steps with breaks in between, the browser can respond in the meantime.
• Load third-party scripts later. Tracking, chat and advertising often run exactly when visitors first click, see third parties.
• Lean HTML. The more elements a page has, the longer every change to the layout takes, see page weight.
• Immediate visible feedback. Show first that the input has arrived, a pressed button for example, and do the heavy work afterwards.

How to improve it

INP is the job of whoever writes or includes the page's scripts, that is your agency or development team. In the "Performance" tab of Chrome's developer tools they can see which interaction is slow and which script runs during it. Whatever is included through the tag manager is for whoever manages it to review.