ImageCrush

Image prep is a constant part of our client work. Websites need hero images resized, cropped to set aspect ratios and converted to WebP or AVIF. Game and AR projects need textures and marketing assets exported at exact sizes. The Mac tools we used each covered one piece, so a single batch often went through three apps and a lot of manual renaming.

We wanted one tool that would take a folder of mixed images and crop, resize, convert, compress and rename it in a single pass. It had to run entirely on the Mac, with nothing uploaded to a server, and sell at a fair one-time price. Nothing on the market did all of that, so we built it.

We soon saw that web developers, designers and photographers had the same problem. That raised a harder question: could a small studio build a polished app and also handle what a product company handles, from App Store submission and licensing to a website and search visibility? An app doesn't sell just because it exists. It has to be found.

One product from day one

We treated the app and its website as one product. The website's repository was created the day after the app's first commit, and while the app was being built we researched the market, studied competitors and drafted the App Store copy. As features shipped in the app, they showed up on the site.

A native Mac app, built fast

Working from a detailed product spec, we built ImageCrush in Swift and SwiftUI. It reads and writes JPG, PNG, WebP, HEIC and AVIF, runs batches concurrently with pause and resume, and feels like part of macOS, with drag and drop, undo and full VoiceOver support. Preset sequences turn one drag into a full-size WebP, a thumbnail and a social crop, and a per-image crop editor with 17 aspect-ratio presets keeps every subject framed. Under the hood, we added the AVIF encoding macOS doesn't offer natively and moved WebP to libwebp's multithreaded encoder, cutting a 4 MB photo from about 13 seconds to about 2.

One codebase, two storefronts

ImageCrush sells directly, with a 7-day free trial, and on the Mac App Store. One codebase produces both builds. The direct version adds licensing, the trial and automatic updates through Sparkle, and is signed and notarized by Apple. The App Store edition follows Apple's sandbox and review rules. A documented release process takes each new version to both storefronts the same way every time.

A website built to be found

We designed imagecrush.io in Figma and built it in Next.js to turn searches into trials. A product homepage with short feature demos sits alongside six landing pages for tasks people actually search for, such as HEIC to JPG, batch resizing and WebP conversion, all on one reusable template. A blog of long-form guides, honest comparisons and original research builds authority. Structured data, an llms.txt file and automatic IndexNow submissions help it show up in Google and in AI assistants, and listings on AlternativeTo, SaaSHub and MacUpdate put it where Mac users compare tools.

ImageCrush is a shipping, paid product, and it's part of our daily workflow: we use it to optimize the images on every website we build, for clients and for Rocket 5 Studios. On the website side, one lesson stood out. Original work earns attention. Our watermarking research, built from 324 removal attempts across four AI models, quickly became the site's most-visited page, and it's the page AI search features pick up most often.

• Speed to market: first commit to version 1.0 in 19 days
• WebP encode time: about 13 seconds → about 2 seconds for a 4 MB photo
• Distribution: direct sales and the Mac App Store from one codebase
• Website: 6 landing pages and 7 long-form articles, ready for search and AI answers

Building ImageCrush took us through every part of a modern app launch: App Store rules, code signing and notarization, licensing, updates, and visibility in search and AI answers. That experience now goes into our app development work for clients: the app, the website that markets it and the plan that connects them.

Swift, SwiftUI, ImageIO, libwebp, libavif, Sparkle, Freemius, Next.js, React, TypeScript, Tailwind CSS, MDX, Vercel, Resend, Figma