Google Just Made LCP Harder. Here's What That Costs You.

Google changed the rules in March 2026, and most stores don't know yet

On March 18, 2026, Google confirmed in a Search Central blog post that INP (Interaction to Next Paint) is now an equal ranking signal alongside LCP and CLS. At the same time, the "Good" LCP threshold dropped from 2.5 seconds to 2.0 seconds.

That second change is the one most stores will feel first. If your LCP was sitting at 2.3 seconds, you were passing before. You are not passing now. You are in the "Needs Improvement" zone, and IdeaFueled's analysis of the March update found that sites with INP above 200ms saw average position drops of 0.8 places following the core update.

A 0.8-place drop sounds small. On a competitive search term, it can move you off the first visible row on mobile. And most Shopify traffic is mobile.

What LCP actually measures, and why 2.0 seconds is harder than it sounds

LCP measures when the largest visible element on a page, usually a hero image or a product photo, finishes loading. The 2.5-second threshold already caught a lot of stores out. Moving it to 2.0 seconds catches more, because the gap between "fine" and "failing" is now paper thin.

The average Shopify store runs 15 to 25 third-party scripts. Review widgets, loyalty apps, live chat, upsell tools, analytics tags. Each one adds to your page weight and blocks the main thread while it loads. A script that adds 300ms to your load time might not have mattered at the old threshold. At 2.0 seconds, it might be the thing that tips you over.

Portent's analysis of over 100 million page views found that a site loading in 1 second has an e-commerce conversion rate 2.5 times higher than a site loading in 5 seconds. The relationship is not linear, most of the conversion damage happens in the first few seconds, which is exactly the range where the LCP threshold change bites.

The revenue maths

The numbers from the IdeaFueled piece are worth sitting with. A one-second delay in page load reduces conversions by 7%. For a store doing £100,000 a month, that single second costs £84,000 a year.

That is not a developer problem. That is a commercial problem.

Google's own case study compendium documents Vodafone Italy improving LCP by 31% and achieving 8% more sales. Cdiscount improved all three Core Web Vitals metrics and attributed a 6% revenue uplift during their Black Friday sale to that work. These are not theoretical gains, they are measured outcomes from real traffic.

The Rakuten 24 A/B test cited in the IdeaFueled analysis is even more striking: good LCP scores drove a 53.37% increase in revenue per visitor and a 33.13% increase in conversion rate, with CWV optimisation as the only variable between the two groups.

Shopify's own research found that even a half-second improvement in site speed produces measurable conversion lift. Half a second. Most stores are giving away far more than that through unaudited third-party scripts and unoptimised images.

INP is now a primary ranking signal, and it's different from LCP

INP measures responsiveness, how quickly the page reacts after a user taps, clicks, or types. The threshold for "Good" is under 200ms. Above 200ms, you're now carrying a primary ranking risk, not just a UX note.

INP tends to spike on pages with heavy JavaScript execution. Product pages with real-time inventory checks, variant selectors that re-render the DOM on every click, review widgets that fire analytics calls on interaction, all of these eat into INP. The problem is usually not a single culprit. It's ten scripts each taking 20ms, adding up to something that visibly lags.

Shopify notes that the likelihood of a bounce nearly triples when a page takes more than three seconds to load. INP failures often do not slow the initial load, they make the page feel broken after it loads, which produces exactly the same result: the user leaves.

Where to start: a practical audit checklist

Before you touch any code, you need to know what your numbers actually are against the new thresholds. Here is a working checklist:

1. Measure your field LCP at p75. Google ranks on field data, not lab data. Use Chrome User Experience Report (CrUX) or PageSpeed Insights to get your 75th percentile LCP. The new threshold is 2.0 seconds. Lab scores from Lighthouse are useful for diagnosis but not what Google uses to rank you.
2. Check INP on your product pages and cart. These are the highest-interaction pages on a Shopify store. If INP is above 200ms on either, that is where to start.
3. Audit every third-party script. List every JavaScript file loading on your storefront. For each one, record the blocking time and transfer size. Most stores have scripts loading that no longer serve an active purpose, abandoned A/B tests, apps that were uninstalled but whose code was never removed from the theme.
4. Check your LCP element is not lazy-loaded. A common mistake is setting loading="lazy" on the hero image, which is almost always the LCP element. That tells the browser to deprioritise loading the exact asset that determines your score.
5. Look at your CLS on mobile. Cumulative Layout Shift is often caused by images without explicit dimensions, or by third-party banners that load after the initial render and push content down. This is particularly disruptive on mobile where the viewport is narrow.
6. Compress and serve modern image formats. WebP is now baseline support across all major browsers. If your product images are still serving JPEG at full resolution, that is easy weight to cut.
7. Defer non-critical scripts. Review widgets, loyalty popups, and live chat tools do not need to block the main thread on page load. Adding defer or async attributes, or loading them after the initial paint, is often the single highest-impact change available.

The script problem most Shopify stores ignore

The audit checklist above is easier to action if you can actually see which scripts are responsible for your blocking time. Most merchants cannot. They know they have a review app, a loyalty app, a upsell tool, and a live chat widget, but they have no idea which one is contributing 400ms of blocking time and which one is negligible.

This is one of the things Bloodhound surfaces directly. The script analytics feature identifies every JavaScript file on your storefront and ranks it by performance impact, blocking time and transfer size. When you are trying to recover LCP after a threshold change, knowing which script to remove or defer first is not a small thing. It is the difference between spending two hours on the right fix and spending two weeks guessing.

Bloodhound also monitors your Core Web Vitals continuously against Google's thresholds, so if a new app install pushes your LCP above 2.0 seconds, you know about it the same day rather than when you notice traffic dropping a month later.

This is not a one-time fix

The March 2026 update changed two things permanently: the LCP threshold is lower, and INP is now a primary signal. But core updates happen multiple times a year, and Google has been consistently moving towards stricter performance requirements since 2021. Stores that treat this as a one-time audit and then walk away will find themselves back in the same position the next time thresholds shift.

The stores that benefit most from performance investment are the ones that monitor continuously, catch regressions when they are introduced rather than when they are already costing traffic, and have enough visibility into their script inventory to make fast decisions when something new gets installed.

The new LCP threshold is 2.0 seconds. Check your field data today.

Sources

• Portent — "A site that loads in 1 second has an e-commerce conversion rate 2.5x higher than a site that loads in 5 seconds."
• Shopify — "Even a half-second improvement in site speed can increase your conversion rate."
• web.dev (Google) — "Vodafone (Italy) improved LCP by 31% to achieve 8% more sales."
• web.dev (Google) — "Cdiscount improved all 3 metrics which contributed to 6% revenue uplift in their Black Friday sale."
• Shopify — "The likelihood of a bounce nearly triples if a page takes more than three seconds to load."