Why image optimization matters

Images are usually the largest part of a web page. On a typical content site, they account for 50 to 70% of total page weight. A single unoptimized hero image can weigh more than the entire rest of the page combined, including HTML, CSS, JavaScript, and fonts.

That weight has real consequences. Pages load more slowly, especially on mobile connections. Google's Core Web Vitals penalize slow Largest Contentful Paint (LCP), which is usually an image. Every second of delay reduces conversions measurably. And if you are serving images through a CDN or hosting provider that charges by bandwidth, oversized images cost money.

The good news: image optimization is one of the few performance improvements that is fully under your control, costs nothing, and takes minutes with the right tools.

Step 1: Resize to the dimensions you actually display

This is the single most impactful optimization, and it is the one most people skip. A modern phone camera produces images at 4000 to 6000 pixels wide. If your website displays them at 800 pixels wide, you are shipping five to seven times more pixels than the browser will ever use. The browser downloads the full file, then throws away 90% of the data.

Resizing to the maximum display width cuts file size by 70 to 95% instantly, with zero visible quality loss. For responsive layouts, generate two or three sizes (for example, 800px, 1200px, and 1920px) and serve the appropriate one via srcset. For most sites, a 1200px or 1920px cap covers the majority of use cases.

Resizing should always be a downscale, never an upscale. Blowing a 600px image up to 1200px adds no real detail. It just adds blur and file size. If you need larger dimensions, start from a higher-resolution original.

You can batch-resize an entire folder at once using the BulkImgResize tool. Set a target width, drop in your images, and download the resized set as a ZIP.

Step 2: Choose the right format

Format matters more than most people realize. At equivalent visual quality, different formats produce wildly different file sizes.

WebP is the best default in 2026. It matches or beats JPEG quality at 25 to 34% smaller file sizes, and matches or beats PNG quality at 26% smaller file sizes while supporting transparency. Every current major browser supports it.

JPEG is still fine as a fallback, but there is no reason to use it as your primary format for new images. It cannot handle transparency, and its compression algorithm is older and less efficient than WebP.

PNG should only be used when you need pixel-perfect lossless output, such as UI screenshots, diagrams with sharp text, or images you will edit again later. For web delivery of graphics, lossless WebP is a direct replacement at 26% smaller size.

AVIF offers even better compression than WebP in some cases, but browser support is less consistent and encoding is slower. It is a promising format to watch, but not yet the default choice for most sites.

For bulk format conversion, use the WebP converter. For mixed JPEG workloads, see the guide on compressing multiple JPGs.

Step 3: Compress at a sensible quality level

Both JPEG and WebP are lossy formats, meaning they discard some data to achieve large size reductions. The quality setting controls how aggressive that discard is. At 100%, the file is large and the quality is near-original. At 50%, the file is small but visible artifacts appear.

For photographs, the practical range is 70 to 85%. At 85%, most people cannot tell the difference from the original at normal viewing sizes. At 70%, file sizes drop substantially with only a subtle difference that is hard to see without zooming in. Below 60%, artifacts become visible in smooth gradients like skies and in fine skin texture.

For graphics, screenshots, and images with flat colors or text, use lossless formats (PNG or lossless WebP). Lossy compression introduces visible blocky artifacts around edges in these cases.

Step 4: Strip metadata

Most camera images carry EXIF metadata: GPS coordinates, camera model, lens, exposure settings, timestamps, and sometimes a thumbnail. This data can add 10 to 50 KB per image. For a page with twenty images, that is up to 1 MB of useless payload.

There is also a privacy dimension. GPS coordinates in EXIF can reveal where a photo was taken, which is a real risk for photos taken at home, at client sites, or anywhere sensitive. Stripping EXIF is both a performance win and a privacy win.

The good news: any workflow that re-encodes through a modern image processor (including the browser's Canvas API) strips EXIF automatically. You do not need a separate tool. Resizing and compressing already remove it.

The order matters

Do these four steps in order: resize first, then choose format, then compress, then strip metadata (which happens automatically). The reason is that resizing before compression gives the compressor far less work to do and produces smaller output. Compressing a 4000px image and then resizing it throws away most of the compression benefit.

A good workflow does all four in a single pass: pick a target width, choose an output format, set a quality level, and the tool handles resizing, conversion, compression, and EXIF stripping together. That is exactly what BulkImgResize does, and it is why the tool is designed around a single "process batch" operation rather than four separate steps.

Core Web Vitals: the "so what"

Google's Core Web Vitals measure the real-world experience of loading and interacting with a page. They are a direct ranking factor, and they are the reason image optimization has become impossible to ignore for anyone who cares about search.

Largest Contentful Paint (LCP) measures how long it takes for the largest visible element (usually a hero image) to render. Oversized images are the most common cause of poor LCP scores. Optimizing image size, format, and dimensions is usually the single biggest LCP improvement you can make.

Cumulative Layout Shift (CLS) measures visual instability while the page loads. Images without explicit width and height attributes cause layout shifts when they finally render. Always set explicit dimensions on your <img> tags so the browser can reserve space before the file arrives.

Interaction to Next Paint (INP) is less directly tied to images, but oversized images do consume memory and can slow down the browser on lower-end devices, which affects interactivity after the page loads.

A simple checklist

  • ✓ Resize every image to its maximum display width (or two to three sizes for responsive layouts)
  • ✓ Use WebP for almost everything; JPEG and PNG only where required
  • ✓ Compress photographs to 70 to 85% quality
  • ✓ Use lossless WebP or PNG for screenshots, icons, and graphics with text
  • ✓ Strip EXIF data (happens automatically when re-encoding)
  • ✓ Set explicit width and height attributes on <img> tags
  • ✓ Use srcset and sizes to serve the right size per viewport
  • ✓ Add loading="lazy" to images below the fold
  • ✓ Never lazy-load your LCP hero image (that hurts LCP)
  • ✓ Serve images from a CDN with proper cache headers

How much weight can you actually save?

The combined effect is dramatic. A typical camera JPEG at 4000px wide, 6 MB, optimized properly, becomes:

  • Resize to 1200px: about 600 KB (90% reduction)
  • Convert to WebP at 80% quality: about 120 KB (80% further reduction)
  • EXIF stripped automatically: about 100 KB final (another 15 to 20% reduction)

That is roughly a 60x reduction from 6 MB to 100 KB, with no visible quality difference at the display size. This is the difference between a 6-second load and a 0.5-second load on a slow connection. It is the difference between a failed LCP score and a passing one.