INP is the Core Web Vital nobody is talking about (and probably the one quietly tanking your rankings)

Here's a stat that surprised me when I dug into it: across the small business sites I audit, INP is the Core Web Vital that fails most often. By a wide margin. LCP usually passes. CLS usually passes. INP fails on roughly six out of ten sites, often badly.

And nobody talks about it. The SEO industry is still optimizing for FID, the metric Google deprecated in March 2024.

What INP actually measures

INP — Interaction to Next Paint — measures the latency of every user interaction on a page (clicks, taps, key presses) and reports the worst one (technically the 98th percentile for high-traffic pages). The threshold for "good" is 200 milliseconds. Above 500ms is "poor."

FID, the metric INP replaced, only measured the very first interaction. That made it easy to game — sites would defer all heavy JavaScript past the initial click and still pass FID even if every subsequent interaction was a slideshow. INP closes that loophole. Every interaction counts. The slowest one defines your score.

Why so many sites fail it

The shift from FID to INP exposed a category of performance problems that everyone had been ignoring: post-load JavaScript bloat. The page loads fast, the first click responds fast, and then the user clicks a navigation menu that triggers 800ms of React state recalculation, and INP scores it accordingly.

The biggest offenders I see, in rough order of frequency:

• Large React or Vue trees that re-render aggressively on interaction. Common in dashboards, e-commerce filters, and anything with controlled inputs that update on every keystroke.
• Third-party widget initialization. Live chat widgets, analytics platforms, A/B testing tools, ad networks. They wait until idle to set up, then execute heavy work on the next user interaction.
• CSS that triggers layout on hover. Animations using width or height instead of transform force layout recalculation. Multiply by every nav item the user hovers over.
• Synchronous storage access. Reading from localStorage during a click handler is fast on a fast machine and devastating on a slow one. INP is measured at the user's machine, not yours.

How to actually diagnose it

The challenge with INP is that lab tests don't catch it. Lighthouse runs an automated load and reports the closest equivalent metric, but it doesn't simulate real user clicks across a session. You need field data.

Field data means PageSpeed Insights' "Field Data" tab, Search Console's Core Web Vitals report, or any RUM tool. The lag time is the hard part — field data updates on a 28-day rolling window, so a fix you ship today shows up in CrUX data a month from now. Plan accordingly.

For diagnosing the specific interactions that are slow, Chrome's "Performance" tab now has an "Interactions" track that shows every user interaction and the rendering work that followed. Trigger the interactions you suspect (open the menu, type in the search box, click the filter), then look for tasks longer than 50ms attached to those interactions. That's your culprit.

The fix that helps the most sites

Audit your third-party scripts. Disable each one temporarily and re-test INP. Most sites have 4-8 third-party scripts and removing the worst 1-2 typically drops INP by 200-400ms. Whether you can actually remove them depends on the business, but you'll at least have leverage in the conversation about whether the live chat widget is worth the ranking cost.

Beyond third-party scripts, the highest-impact fixes are: debouncing search input handlers, virtualizing long lists, breaking up long JavaScript tasks with scheduler.postTask() or setTimeout(fn, 0), and migrating CSS animations to transform and opacity.

Why this matters for SEO specifically

INP is part of the page experience signal Google uses for ranking. Failing INP doesn't single-handedly tank a site, but in close ranking matchups, it's the kind of thing that decides which page wins. And because almost nobody is optimizing for INP yet, fixing yours is one of the few SEO improvements where you can actually leapfrog competitors who are spending all their time on backlinks.

Run your site through the Core Web Vitals Checker and look specifically at the INP score. If it's above 200ms, you have the most actionable performance work in your queue. If it's above 500ms, you have an active ranking problem.