Google’s Mueller Flags A Case On Why LCP Fixes Miss the Target

Google Search Advocate John Mueller highlights a caller lawsuit study explaining why definite Largest Contentful Paint improvements often neglect to nutrient results. In a layout that varies by store, the browser whitethorn attraction connected the incorrect element, causing each consequent optimizations to target thing that was ne'er the LCP successful the first place.

The case study was published connected web.dev connected June 24, and it specifications a twelvemonth of Core Web Vitals activity astatine the ecommerce level Nuvemshop. Initially, the squad suspected image weight aliases server latency arsenic the main issues. However, their study pointed them toward really the browser selected the LCP element. After optimizing constituent selection, image priority, lazy loading, and separator caching, Nuvemshop reports that the stock of its stores pinch bully LCP scores accrued from 57% to 96%.

How the Wrong Element Gets Picked

On Nuvemshop, merchants tin shape their homepage sections successful immoderate bid they prefer, resulting successful carousels, banners, and merchandise grids appearing successful various positions crossed different themes. According to the lawsuit study, carousels were coming connected 85% of storefronts, and this is wherever the measurement encountered an issue.

The rumor stemmed from CSS transitions connected carousels and banners. These transitions delayed erstwhile elements became visible to the browser’s LCP detection. While shoppers saw the carousel load first, the browser occasionally identified a banner further down arsenic the LCP constituent because its visibility was not delayed by a transition. Consequently, according to Nuvemshop, the erstwhile optimization efforts focused connected elements that were ne'er the LCP.

The investigation identified 3 main causes. Transitions were pushing discovery to the incorrect frame, lazy loading was delaying images supra the fold that needed to load immediately, and the captious images lacked privilege signals. Additionally, measurements originated from some class and merchandise pages, not only homepages, truthful the solution had to beryllium effective crossed different page types.

The Three LCP Fixes

The lawsuit study includes a 4th change, separator caching to trim latency. However, the halfway element-detection adjustments progressive 3 modifications successful really the apical of the page renders and loads. Nuvemshop implemented each 3 changes crossed its main themes and page types. These adjustments are straightforward, pursuing communal web capacity champion practices.

The squad removed CSS transitions from the apical sections truthful they show instantly, enabling the browser to admit them arsenic LCP candidates without delay.

They besides pulled loading="lazy" disconnected the first image successful those apical sections. web.dev’s ain guidance is blunt connected this, warning against lazy-loading the LCP image because it ever adds load delay.

The past portion was fetchpriority="high" connected the apt LCP image, which tells the browser’s preload scanner to drawback it sooner. Nuvemshop says it wrapped that successful validation logic truthful the awesome fires only wherever the constituent could really beryllium the LCP candidate. Google’s proposal is to reserve the property for one aliases 2 cardinal images, since if everything is precocious priority, thing is.

What Else Nuvemshop Reports

Nuvemshop reports that its wide Core Web Vitals walk complaint improved from 48% to 72% passim the year, not conscionable the LCP figure.

The institution reviewed the aforesaid Brazilian stores progressive successful some January 2025 and 2026. For mobile visitors arriving via Google integrated search, it reports an 8.9% summation successful conversion complaint and an 8.4% emergence successful cart engagement.

Keep successful mind these are self-reported figures, and the shopping numbers are a year-over-year comparison of that cohort, not a controlled test. Nuvemshop ties the guidance to its velocity activity by leaning connected Deloitte investigation commissioned by Google, which recovered that shaving 0.1 seconds disconnected load clip tin raise unit conversion by 8.4%.

Why This Matters

Before compressing different leader image, double-check that you’re optimizing the constituent the browser really treats arsenic LCP. This is particularly important successful template-driven aliases carousel-heavy layouts, wherever it mightiness not hold.

Our ain caller sum lands successful the aforesaid spot. In a look astatine Core Web Vitals crossed CMS platforms, HTTP Archive information showed real-world LCP tends to break erstwhile a level is slow to fto the browser observe the main image, aliases erstwhile it forces that image to compete pinch excessively galore different high-priority resources, not conscionable erstwhile a page is heavy. Barry Pollard has besides walked done how to trace an LCP problem successful PageSpeed Insights earlier you touch immoderate code.

The conversion nexus reflects Nuvemshop’s mentation of the data. Core Web Vitals are only a insignificant ranking factor, and Google has emphasized relevance arsenic much important, truthful a faster shop is amended publication arsenic a shopping acquisition summation than a ranking lever.

Looking Ahead

This is conscionable 1 company’s self-reported experience, truthful the shopping gains are champion publication arsenic directional. The method itself isn’t caller either, since web.dev has been advocating for this benignant of discovery-and-priority activity for years. What the lawsuit study adds is simply a worked illustration of identifying the existent LCP constituent crossed layouts that alteration from 1 shop to another, which is the difficult portion connected immoderate customizable platform.

Featured Image: victoriatudosi/Shutterstock