INP (Interaction to Next Paint)

It is the newest of the three Core Web Vitals metrics; in March 2024, it replaced FID (First Input Delay). Unlike FID, it does not assess only the first click, but responsiveness throughout the entire visit, with one of the slowest interactions factored into the result. Google considers INP up to 200 ms good; above 500 ms is poor.

For an online store, it is the metric customers feel most: they tap a filter, size, or add-to-cart button and nothing happens for a moment. In our measurement of 85 of the largest Czech online stores (October 2026), INP was the most common reason a website failed Core Web Vitals - it happened to 56 of the 58 online stores that failed. The median was 244 ms, above the limit. Tenths of a second sound insignificant, but customers perceive them as hesitation - they tap again, choose another option, or leave.

JavaScript that blocks the browser’s main thread almost always slows responsiveness: while a script runs, the page cannot respond. It helps to send less JavaScript (server-side rendering instead of rendering in the browser), split long tasks into shorter ones, defer third-party scripts such as chats and analytics code, and show a response immediately after a click before doing the heavy work. That is why the form with the calculator on this website is downloaded and hydrated only when a visitor scrolls to it or starts using it.

INP cannot be measured in a lab; Lighthouse shows Total Blocking Time instead, which is closely related to it. Actual INP exists only in data from real visitors (Chrome UX Report, PageSpeed Insights), so that is where we read it in an audit.