What is Interaction to Next Paint (INP)?

Interaction to Next Paint (INP) measures how long a page takes to visibly respond after someone taps, clicks or types, judged across the whole visit rather than the first interaction.

Want to know where your site stands on INP?

I’ll test your site on a real mid-range phone profile and send you the two or three fixes that matter most, free.

No newsletter. Just what you asked for, from me. Privacy

In plain English

Every time a visitor opens a menu, ticks a checkbox or presses a button, the browser has to run code and then redraw the screen. INP records those delays and reports roughly the slowest one. If the main thread is busy with scripts, the tap just sits there.

It became a Core Web Vital in March 2024, replacing First Input Delay. Google rates 200 milliseconds or less as good and anything over 500 milliseconds as poor.

Why it matters for a small business

A lag after a tap feels like a broken site. People tap again, open the wrong thing, or give up on a form halfway through. For a business that relies on enquiries, a sluggish mobile menu or quote form costs leads quietly, because nobody emails to say the button felt slow.

An example

An Airdrie renovation contractor’s site loads three analytics tools, a heatmap script and a review carousel. Tapping the mobile menu takes most of a second to respond on a mid-range Android phone. Dropping the unused heatmap and deferring the carousel until it scrolls into view brings INP comfortably under 200 ms.

Common mistakes

Where I see INP go wrong most often.

  1. Judging responsiveness on a fast laptop, where heavy scripts barely register.

  2. Adding tracking and marketing tags one at a time without ever reviewing the total.

  3. Assuming a good load speed means good INP, when the two measure different moments.

FAQ

INP, answered

  1. Google considers 200 milliseconds or less good, 200 to 500 milliseconds needs improvement, and over 500 milliseconds poor, at the 75th percentile of real visits.

  2. INP needs real interactions, so it only appears in field data. Lighthouse uses Total Blocking Time as a lab stand-in, which points in the same direction.

Keep reading

Let’s begin

Let’s build something you’re proud to share.

Let’s Talk