INP measures how long the page takes to react to a tap or a click, from the moment you press to the moment something changes on screen. If the page shows up fast but the menu opens late or the button “feels dead”, INP is the problem. It is almost never the server: it is too much JavaScript keeping the browser busy, and that runs on the visitor’s phone, not on the machine hosting the site.
What INP catches, and what it does not
INP looks at real interactions: taps, clicks and key presses. It does not look at scrolling or at load time. That is why a site with a good LCP can have a poor INP, and the other way round. It replaced the old FID and is stricter, because it counts every interaction in the visit and not only the first.
The usual causes
| The cause |
What you see |
What to do |
| Third-party scripts |
Chat, ads, tracking pixels, maps, embedded videos. |
Remove the ones nobody uses; load the rest only when the person asks for them (for example, chat only when they click its button). |
| Many plugins and page builders |
Each one adds JavaScript that runs at start-up. |
Switch them off one at a time and measure. See updating themes and plugins without breaking anything. |
| Heavy menus, carousels and animations |
The tap on the menu lags, or the animation stutters. |
Simplify, or swap for a version that uses only CSS. |
| Scripts that run everything at once |
The browser stays busy for one long block and ignores taps. |
Defer whatever is not needed at the start (defer), and split long tasks if you have someone who codes. |
Measuring and fixing it, step by step
| 1 |
Measure with real data. The field-data panel in PageSpeed Insights and the Core Web Vitals report in Search Console show INP from real visitors. The lab test does not measure INP the same way, because nobody clicks for it.
|
|
| 2 |
Reproduce the problem. In Chrome, F12, Performance panel: record and click the menu or button that feels slow. “Long tasks” are marked in red.
|
|
| 3 |
See who is taking the time. In the activity list, the script file name shows whether it belongs to the theme, a plugin or an outside service.
|
|
| 5 |
Measure again, preferably on a plain phone and not on your powerful computer. That is where the difference is felt.
|
|
|
The server does not fix INP. A bigger plan does not make a visitor’s phone run JavaScript any faster. If your INP is poor, changing plan is not the answer; taking weight off the browser side is.
|
|
Test on a modest phone. Your computer hides the problem. If the page reacts well on an old phone, it reacts well almost everywhere. The limits Google counts as good or poor are on web.dev and they change: check them there.
|
|
Found the script that slows the interactions and not sure whether you can remove it? Send us the address and the file name.
Open a support ticket
|
RECOMMENDED PRODUCT Web hosting with cPanel Domain and SSL included, daily backups and the panel you already know. from 5.940,00 Kz/mo (3-year plan, with coupon) See plans |
Português
English
