The Sizes That Actually Matter
“Bulk image resizer” is only useful if it spares you the lookup table. These are the numbers the major platforms publish or reward as of 2026 — the presets above encode them so one click does the math:
| Platform | Min for zoom | Recommended | Gotcha |
|---|---|---|---|
| Amazon | 1000 px long edge | 2000 × 2000 px | Below 1000 px: no pinch-zoom, listing looks broken |
| Shopify | None hard | 2048 × 2048 px | Storefronts serve CDN renditions — 2048 covers retina grids |
| Etsy | 500 px | 2000 px long edge | First photo becomes the thumbnail on search |
| eBay | 500 px (or flagged) | 1600 × 1600 px | Images below the minimum are demoted in results |
| Walmart | 500 px | 2200 × 2200 px | Promo text, watermarks or borders = rejection |
Why “Fit Inside” Beats Forced Dimensions
Most desktop batch tools offer one number: width × height, applied to every file. Camera catalogs are a mix of landscape and portrait shots, so a forced 2000 × 2000 on a rectangular photo either squashes the product or — in better tools — center-crops it. Cropping is silent data loss: your product touches the frame edge and Amazon-style white-margin checks get nervous. The fit-inside box scales each image to the largest size that still fits inside 2000 × 2000 without touching the aspect ratio, which is why every preset here uses it. If your catalog really is one shared ratio, exact mode is still there.
Resizing Without the Upload Risk
Online bulk resizers normally mean shipping your unreleased catalog to someone’s server, queueing it, and hoping the “files auto-delete in 24 hours” promise is true. BatchBG does the pixel math with the same Canvas API the browser uses to render this page: the file is decoded, scaled, re-encoded and zipped inside your own tab. Nothing traverses a network, so nothing can leak, be retained, or train anything. The same no-upload design runs through the rest of the toolkit — the bulk background remover, the batch image compressor, and the offline editor — so a resize → trim-edges → compress pipeline never puts one file online.
A Full Catalog Workflow
- Start from originals. Resize copies, not your archival files — downscaling throws pixels away permanently.
- One pipeline, three steps. Remove backgrounds first if you need white, then resize to platform dimensions, then compress to the file-size ceiling. Order matters: background removal on 6000 px originals keeps edge detail.
- Keep zoom eligibility. If a marketplace promises zoom, never let the long edge fall under 1000 px — this 1000 × 1000 batch page locks exactly that floor, or see the Amazon image spec and Shopify size guide for the full rules, or jump straight to resizing a folder to 2000 × 2000, the size that satisfies most of them at once, or to 1024 × 1024 for app-icon masters and AI reference folders, or to 1500 × 1500, the size Google Merchant Center itself recommends for product feeds, or to 2048 × 2048, Shopify's stated display-best square that maps 1:1 onto the grande variant, or to 512 × 512 for Google Play icons, WhatsApp stickers, PWA manifests and Stable Diffusion 1.5 inputs, or to 800 × 800 for eBay zoom activation and Google Shopping's minimum — the power-of-two chain from 512 to 2048 covers both icon and marketplace sizes.
- Batch by marketplace, not by product. Resizing your whole catalog to one platform’s size and then again for the next doubles work; run one fit-inside 2048 master batch, then derive platform copies.
- Name stays intact. Outputs keep original filenames inside the ZIP, so listing CSVs that reference image names survive the round-trip — one reason sellers keep this tab over remove.bg-style alternatives.