Sharjeel TahirYour everyday toolkit
Menu
PRODUCTIVITY

Image optimization workflow: fast sites, sharp photos

Every image you put online is a trade-off between quality and size — and most people get it wrong in both directions: bloated originals that load slowly, or over-compressed mush. Here’s a repeatable workflow that gets it right every time, for websites, WhatsApp, and print.

Published October 9, 20269 min readBy Sharjeel Tahir

Why does image optimization matter so much?

Images are typically most of a web page’s weight — often sixty to eighty percent. A single unoptimized phone photo can outweigh an entire article’s text, code and styling combined, doubling load time. On mobile networks with limited data, that is the difference between a visitor who stays and one who leaves before the page renders.

It’s not just websites. Oversized images clog WhatsApp chats, exceed email attachment limits, get rejected by upload portals, and fill phone storage. And the fix is almost always the same sequence of operations — which is why having a workflow beats having a bag of tricks.

The good news: modern compression is extraordinarily good. A photo compressed to 10–15% of its original size is typically indistinguishable from the original on screen. The visible-quality loss people fear from compression almost always comes from doing it wrong — wrong format, wrong dimensions, or compressing an already-compressed file repeatedly.

Step 1: how do you choose the right image format?

Format choice is the highest-leverage decision in the workflow. Photos: JPG for universal compatibility, WebP for web use (25–35% smaller than JPG at equal quality). Graphics with transparency or sharp edges (logos, icons, screenshots with text): PNG. iPhone photos: HEIC is efficient but incompatible with much of the world — convert to JPG for sharing.

Our image converter handles all of these transitions in the browser. The common mistakes: PNG for photographs (a photo saved as PNG can be 5–10x larger than the same photo as JPG, for zero visible benefit), and JPG for logos with transparency (JPG doesn’t support transparency, so you get ugly white boxes). Match the format to the content type and you’ve won half the battle before compressing anything.

For web publishing specifically, WebP is the sweet spot today: broad browser support, excellent compression, transparency support. Convert with a JPG to WebP converter as the last step before upload. (AVIF compresses even further but has spottier tooling support — check your pipeline before committing to it.)

  • Photographs → JPG (universal) or WebP (web)
  • Logos, icons, text graphics → PNG
  • iPhone HEIC → convert to JPG before sharing
  • Never PNG for photos; never JPG for transparency

Step 2: why resize before you compress?

A 4000-pixel-wide phone photo displayed at 800 pixels is carrying 25x more pixels than the screen can show — and every one of them costs bytes. Resizing to the display size first is the single biggest size reduction available, often 80–90% on its own.

Always resize before compressing: compressing first and resizing after wastes the compressor’s effort on pixels you’re about to throw away. Know your target dimensions. Blog content images: 800–1200px wide is plenty. Hero/banner images: 1600–2000px. Thumbnails: 300–400px. WhatsApp sharing: 1280px on the long edge looks sharp on phones while staying small. An image resizer handles exact pixels or percentage scaling in seconds.

One subtlety: resize from the original, not from an already-compressed copy. Each generation of compression discards detail permanently; starting the resize from the highest-quality source you have keeps the final result clean. Keep originals archived, and treat every resized/compressed copy as disposable.

Step 3: how do you compress without visible quality loss?

Compression quality is a slider, and the correct setting is “as low as you can go before you can see the difference.” For JPG photos, quality 75–85% is the sweet spot: below 70%, artefacts appear in skies and gradients; above 90%, file size balloons for imperceptible gains. A good image compressor lets you preview the trade-off.

When you need an exact size — a portal demanding “under 100KB” or “under 20KB” — don’t guess at quality sliders. Use a target-size tool that binary-searches the right compression automatically. It’s faster and far more accurate than trial and error, especially for ID photos and signatures with strict limits.

Check results at 100% zoom on the actual content that matters: faces, text, fine detail. Thumbnails lie — everything looks fine small. And never compress a JPEG twice if you can avoid it: each round re-compresses the artefacts of the last round. One clean pass from the resized original beats three aggressive passes.

  • JPG quality 75–85% for the quality/size sweet spot
  • Use target-size tools for exact KB requirements
  • Verify at 100% zoom, not thumbnail size
  • One clean compression pass from the resized original

Step 4: when should you convert formats as part of the workflow?

Conversion enters the workflow at two points: at the start, getting an incompatible format into something workable, and at the end, getting an image into the delivery format. iPhone HEIC photos are the start-of-workflow case — convert HEIC to JPG first, since most editors and portals don’t accept HEIC.

WebP files downloaded from websites are the reverse problem: perfectly good images in a format some desktop apps and older systems don’t open. WebP to JPG converts them to the universal format. And SVG logos needed as images convert via SVG to PNG at exactly the pixel size required — rendering a vector at the final size beats scaling a raster every time.

Order of operations for the full pipeline: convert to a working format (if needed) → crop/edit → resize to target dimensions → compress → convert to delivery format (if different, e.g. WebP for web). Do it in this order and each step works on the cleanest possible input.

What does the complete workflow look like for common cases?

Blog/website image: start from the original → crop to the composition you want → resize to 1200px wide → compress at JPG 80% (or convert to WebP) → target: 100–200KB. A page of ten such images loads in about a second on 4G instead of ten.

WhatsApp/document sharing: resize to 1280px long edge → compress to under 500KB → send as document (not as photo) if quality matters, since WhatsApp recompresses photos aggressively. For ID documents and forms, compress to the portal’s exact KB limit instead.

Print: different rules — keep 300 DPI at the print size (a 6×4 inch print wants 1800×1200px), use minimal compression (JPG 90%+ or PNG), and never upscale a small image for print; upscaling invents pixels and looks soft. When in doubt, print a small test before committing to the full run.

  • Web: 1200px wide, WebP/JPG 80%, 100–200KB target
  • WhatsApp: 1280px long edge, under 500KB, send as document
  • Print: 300 DPI at final size, minimal compression, test first

Frequently asked questions

What’s the best image format for websites?

WebP: 25–35% smaller than JPG at equal quality, with transparency support and broad browser compatibility. Use JPG as the fallback for maximum compatibility.

How much can I compress a photo without visible loss?

Typically to 10–20% of the original size. JPG quality 75–85% is the sweet spot — below 70% artefacts appear, above 90% wastes bytes for no visible gain.

Should I resize or compress first?

Resize first, always. Resizing a 4000px photo to 1200px removes ~90% of the pixels; compressing first wastes effort on pixels you’re about to discard.

Why is my PNG photo so huge?

PNG is lossless — terrible for photographs, which compress far better as JPG or WebP. A photo saved as PNG can be 5–10x larger than the JPG version with no visible benefit. Convert it.

What size should images be for WhatsApp?

1280px on the long edge, compressed under 500KB. And send important images as documents, not photos — WhatsApp recompresses photos aggressively.

Can I recover quality from an over-compressed image?

No. Compression discards detail permanently — that’s why you keep originals archived and compress copies. Never compress a JPEG multiple times if you can do it once from the source.

Compress your images free, privately

Drop in JPG, PNG or WebP photos and get dramatically smaller files with no visible quality loss — in your browser, no sign-up.

Open the image compressor ↗
Portrait of Sharjeel Tahir
About the author

By Sharjeel Tahir

Sharjeel Tahir is a WordPress and technical SEO specialist based in Lahore, Pakistan. He builds free image tools — compressors, resizers, converters and background removers — and writes practical guides for creators and site owners.

Published: 2026-10-09