Google’s Core Web Vitals thresholds didn’t move in 2026. Your scores might have anyway. Google tightened how it measures Interaction to Next Paint (INP), and stores that were passing comfortably last year are now showing yellow or red in Search Console without changing a single line of code.
If you run a Shopify or WooCommerce store and haven’t checked your field data since the spring, this is the week to do it.
What actually changed
According to a detailed analysis published by WebVitals.tools in April 2026 and updated in June, Google’s 2026 update did not touch the three published thresholds: LCP still needs to come in under 2.5 seconds, INP under 200 milliseconds, and CLS under 0.1. What changed is the measurement methodology. Google tightened how INP captures sustained interaction latency on input-heavy pages, expanded soft-navigation support in the Chrome UX Report (CrUX) for single-page apps, and pushed Time to First Byte (TTFB) higher up the diagnostic stack in PageSpeed Insights, though TTFB still isn’t a ranking signal on its own.
The practical effect, per that same analysis, is that sites which re-measured INP after the methodology change saw an average 18 to 25 percent regression in their 75th-percentile scores compared to their late-2025 baseline. The damage concentrated on listing pages, search results pages, and dashboards, exactly the page types most e-commerce sites lean on.
What to do: Pull your Search Console Core Web Vitals report this week and compare it against a screenshot from before April. If your INP looks worse and you haven’t touched your codebase, this is why.
INP is the metric actually killing your rankings
Of the three Core Web Vitals, INP is the one most stores are failing outright. The same WebVitals.tools analysis puts current origin pass rates at 78 percent for LCP, 84 percent for CLS, and just 72 percent for INP, making it the lowest-passing metric of the three and the newest addition to the ranking signal.
The reason INP is harder than it looks: it measures the slowest interaction across an entire page visit, not just the first click. A product filter that freezes for half a second, a cart drawer that stutters on open, a size-selector dropdown that lags behind a tap. Each one drags your p75 score down, and unlike the old First Input Delay metric, you can’t game INP by having a fast opening interaction and a slow app underneath it.
We saw a version of this dynamic play out earlier this year when a tenth of a second in load time moved retail sales by 8.4%. Interaction speed is turning out to matter just as much as load speed.
What to do: Stop trusting your Lighthouse score alone. Lab tests run on a fast developer laptop and won’t reflect what a customer on a three-year-old Android phone experiences in your cart. Install the `web-vitals` JavaScript library (version 4 or later) to capture real field data, then segment it by device and page template. You’ll usually find the damage concentrated on two or three templates, not the whole site.
Your product filters and cart drawer are the usual suspects
If your INP score dropped, the fix is rarely a full rebuild. It’s almost always a specific interaction with a long-running JavaScript handler behind it. Common culprits on e-commerce sites include filter panels that re-render the entire product grid on every click, cart drawers that recalculate shipping or tax synchronously, and third-party review or chat widgets that attach heavy event listeners to the page.
Framework upgrades alone won’t fix this. React’s concurrent features, `useTransition` and `useDeferredValue`, genuinely help by letting the browser stay responsive during expensive re-renders, but only if a developer actually rewires the slow interaction to use them. Simply bumping a React version number changes nothing.
What to do: Profile your three highest-traffic page templates (usually product listing, product detail, and cart) in Chrome DevTools’ Performance panel. Look for any single task over 50 milliseconds tied to a click or tap. Break it up or defer it before you touch anything else.
There’s a new reason to care: AI search is watching your speed too
Here’s the wrinkle that makes this more than a ranking-signal story. The WebVitals.tools analysis also tracked how often AI answer engines like Google’s AI Overviews, ChatGPT Search, and Perplexity cite fast versus slow sites, and found that sites passing all three Core Web Vitals with a fast TTFB showed a 3 to 4 times higher citation rate than poor-performing sites in the same topical category. The report is upfront that this is a directional, observed pattern from monitored reference sites in Q1 2026, not a confirmed or published ranking factor, so treat it as a strong hint rather than settled fact.
The logic tracks with what we’ve been telling readers preparing for Google’s AI Overviews shopping funnel: fast, well-structured pages are simply easier for retrieval systems to crawl, parse, and trust as a source.
What to do: If you haven’t already, add JSON-LD structured data (Product, Article, or BlogPosting schema depending on the page) with accurate publish dates and authorship. It won’t fix a slow INP score, but it removes one more reason for an AI system to skip past your page in favor of a competitor’s. Our e-commerce SEO checklist and featured snippets guide both cover the structured data basics if you’re starting from zero.
What to fix first if you’re short on engineering time
If your LCP is between 2.5 and 3 seconds, leave it alone for now. That’s a small gap. If your INP is above 200 milliseconds, fix that first. It takes more engineering effort per fix, but you’ll see a bigger ranking and citation impact per hour spent than shaving another few hundred milliseconds off an LCP that’s already close to passing. The one exception: if your LCP is above 4 seconds, that’s severe enough to address before anything else.
The takeaway
Nothing about Core Web Vitals got harder on paper in 2026. The thresholds are the same ones Google published in 2021. What changed is that the measurement finally caught up to how customers actually experience a slow filter or a laggy cart, and a lot of stores that thought they’d solved performance are finding out they hadn’t. Check your INP field data this week. If it’s ugly, you now know why, and you know it’s fixable without a full rebuild.
Sources
- WebVitals.tools, “Core Web Vitals Update 2026: What Changed, What to Fix First,” published April 23, 2026, updated June 27, 2026 — https://webvitals.tools/blog/google-core-web-vitals-update-2026/
- E-CommPartners, “A 0.1s Speed Boost Lifted Retail Sales 8.4%. Here’s How” — https://www.e-commpartners.com/a-0-1s-speed-boost-lifted-retail-sales-8-4-heres-how/
Last reviewed: July 25, 2026
Affiliate disclosure: This site may earn commissions from links in this article at no extra cost to you.
