Performance 14 min read

How to Reduce Image Size for Core Web Vitals

Fix LCP, CLS, and INP issues caused by oversized product imagery — with a store-specific action plan, real before-and-after benchmarks, and the exact settings that move you from Needs Improvement to Good.

This guide is general guidance, not guaranteed outcomes. Image results vary by source image, dimensions, quality setting, and browser encoder — compare visual quality and file size on your own images before publishing. Figures shown are illustrative examples.

Smartphone screen showing a website speed test dashboard with Core Web Vitals metrics

Core Web Vitals are the metrics Google uses to describe whether your page feels fast to a real visitor, and they are part of Google’s page experience signals. For most online stores, images are a significant factor across all three metrics — here’s where they show up, how to fix each one, and a practical action plan you can run.

Google evaluates three Core Web Vitals as part of its page experience signals: Largest Contentful Paint (LCP), Cumulative Layout Shift (CLS), and Interaction to Next Paint (INP). Each has a specific threshold for “Good,” and images touch all three — sometimes directly, sometimes indirectly. Understanding the distinction tells you where to spend your effort.

The three Core Web Vitals at a glance

MetricWhat it measuresGood thresholdHow images affect it
Largest Contentful Paint (LCP)Time until the largest visible element rendersunder 2.5sDirectly — the LCP element is usually an image
Cumulative Layout Shift (CLS)How much visible content shifts during loadunder 0.1Indirectly — missing image dimensions cause shifts
Interaction to Next Paint (INP)Responsiveness to user inputunder 200msIndirectly — heavy images compete for decode time

Largest Contentful Paint (LCP): almost always an image problem

LCP measures how long it takes the largest visible element — on a storefront, that’s often your hero banner or the main product image — to fully render. Google’s “Good” threshold is under 2.5 seconds. On an unoptimized store, a large hero image can take a meaningful share of that budget, which is why LCP is the metric many sellers focus on first.

The fix: shrink the LCP image itself. Convert it to WebP, resize it to the actual rendered width (not the camera-original resolution), and set compression quality around 85–90% since this is the one image on the page where visual quality matters most. Use the WebP compressor’s hero preset to get the right target width for Shopify or WooCommerce automatically.

Secondary fixes that compound with compression:

  • Don’t lazy-load the LCP image. Lazy loading the very image that determines LCP delays it further. Most themes handle this correctly by default, but custom hero sections are a common place this breaks. The first image in the viewport should load eagerly.
  • Use fetchpriority="high" on the LCP image if your theme or platform supports it. This tells the browser to prioritize fetching that image above others on the page.
  • Preload the LCP image with a <link rel="preload"> tag in the document head for an extra head start on the request.
  • Serve the right size. A 2560px hero served to a 375px phone is wasted bytes. Shopify and WooCommerce both generate responsive variants — make sure your theme requests the appropriate size via srcset.

Cumulative Layout Shift (CLS): missing image dimensions

CLS measures how much visible content jumps around as the page loads. The most common store-specific cause is an <img> tag without explicit width/height (or aspect-ratio CSS) — the browser doesn’t know how much space to reserve, so it shifts everything below it down once the image finishes loading. Every product image without dimensions contributes a small shift, and they add up.

The fix: this isn’t solved by compression, but compression makes it easier to solve — smaller files load faster, which shrinks the window in which a layout shift is even visible. Combine faster-loading WebP images with explicit width/height attributes (or CSS aspect-ratio) on every product image, and CLS drops close to zero.

On Shopify, most well-built themes include dimensions automatically. On WooCommerce, check that your theme outputs width/height on image tags and that any custom image blocks (builders like Elementor, Gutenberg image blocks) include them. A common culprit is custom hero sections built with page builders that omit dimensions in favor of CSS sizing — add an explicit aspect-ratio in those cases.

Side-by-side comparison of a slow-loading product page versus a fast, fully loaded product page
The difference between an unoptimized hero image and a properly compressed one shows up directly in your LCP score.

Interaction to Next Paint (INP): indirect, but real

INP measures how responsive your page feels to clicks and taps. Heavy images don’t directly block interaction the way JavaScript does, but oversized images competing for bandwidth and decode time during page load can delay when the main thread becomes free to respond to a shopper’s first tap — especially on mobile devices with slower CPUs decoding large images.

The fix: reducing total image payload (smaller files, fewer bytes to decode) frees up the main thread sooner. This is a secondary benefit of the same WebP conversion work, not a separate task. The same catalog-wide compression pass that fixes LCP also improves INP by reducing the decode work the browser has to do as the page loads.

One additional INP lever worth knowing: lazy-loading below-the-fold images defers their decode entirely until the shopper scrolls to them, which keeps the main thread free during the critical first-paint window. Most modern themes do this automatically via the nativeloading="lazy" attribute — verify it’s present on gallery and thumbnail images.

A store-specific action plan

  1. Run your homepage and a representative product page through PageSpeed Insights and note your current LCP, CLS, and INP scores. PageSpeed Insights reports both lab data (a single simulated test) and field data (aggregated from real Chrome users over 28 days) — field data is what Google uses for ranking, so prioritize it.
  2. Identify the LCP element (PageSpeed Insights labels it directly) — it’s usually your hero image or first product photo.
  3. Compress that image specifically at a high-quality WebP setting (85–90%) using the correct rendered width, and re-upload with the same filename.
  4. Confirm the LCP image is loaded eagerly (not lazy-loaded) and, if possible, marked fetchpriority="high".
  5. Batch-convert the rest of your catalog images to WebP using platform presets — see our batch workflow guide for a repeatable process.
  6. Add explicit width/height or aspect-ratio to any custom image components in your theme that are missing them. Audit with Lighthouse’s “Layout shift elements” report, which lists the specific images causing shifts.
  7. Enable lazy-loading on below-the-fold gallery and thumbnail images if your theme doesn’t already.
  8. Re-run PageSpeed Insights after the changes and confirm the improvement. Field data takes up to 28 days to fully reflect changes, so use lab data to validate immediately.

What kind of improvement to expect

Moving from unoptimized, oversized originals to properly sized, properly compressed WebP can reduce the time it takes the hero or product image to transfer and decode, which can lower LCP. How much depends on the original image size, the new dimensions and quality, the device, and the network — so treat the figures below as an illustrative example, not a guaranteed result.

The illustrative example below shows the kind of change that’s possible when a store replaces oversized JPEG/PNG originals with properly sized WebP and adds explicit image dimensions:

MetricBeforeAfterChange
LCP (mobile, field)3.9s (Needs improvement)1.7s (Good)−56%
CLS0.18 (Needs improvement)0.02 (Good)−89%
INP240ms (Needs improvement)160ms (Good)−33%
Total page weight (product page)4.6 MB1.3 MB−72%
Hero image size1.8 MB (6000px JPEG)240 KB (2560px WebP, 88%)−87%

In this illustrative example, the CLS improvement came from adding explicit dimensions to a custom hero section; the LCP and INP improvements came from the image compression itself. Your own results will depend on your images, theme, and traffic.

Why this matters beyond the scores

Core Web Vitals describe how a page actually feels to load. Faster pages tend to feel more responsive, and broad industry research generally associates faster pages with better engagement — though specific outcomes vary by store, audience, and product, and no particular revenue or conversion result is guaranteed.

Page experience is one of many signals search engines consider. It is best understood as a potential tiebreaker among pages that are otherwise similar in content and relevance — not a guarantee of higher rankings. Improving your Core Web Vitals is one of the few performance levers largely within your control, which makes it worth doing regardless of how search engines weigh it.

Keeping the gains

A one-time optimization pass buys you a score improvement today, but new images keep entering your store. To keep your Core Web Vitals in the “Good” band permanently:

  • Make WebP compression a mandatory step in your product-launch process, before images are uploaded.
  • Set a quarterly reminder to re-run PageSpeed Insights on your homepage and a sample product page.
  • Watch for theme updates that might change image slot sizes or reset lazy-loading settings.
  • If you work with a photographer or VA, give them your settings sheet so images arrive already optimized.

This work compounds with every image on your site, not just the LCP element — want the full breakdown on exactly which quality and resize settings to use per image type? Read best WebP compression settings for websites, or jump straight into the compressor and start with your homepage hero.

Free WebP compressor

Compress your product photos in the next 30 seconds

No sign-up, no upload to a server, no size limits. Drop your JPEGs and PNGs in and get store-ready WebP files back instantly.

Open the compressor