Pic Prune | Apptivate

The problem

The idea for Pic Prune came when the founder realized that he had over 100,000 photos in his library, and more than half were duplicates. He wanted a way to easily compare photos from the same "burst". He had been paying for extra iCloud storage because of this, and he realized that other people were probably in the same boat. There were other utility apps available, but they were badly designed, required a login, or uploaded your photos to a third-party cloud.

The ask was very clear. Pic Prune needed to be beautifully designed, require no login to work, be available in most major languages, and most importantly, no photos could ever leave a user's device.

Oh, and they wanted it built in less than a month. Easy, right?

Pic Prune had a clear idea and no team to build it.

That's where Apptivate comes in.

What we did

We started with a Blueprint: what the app had to prove, who it was for, and what features the first version would leave out. We worked closely with the Pic Prune team on the design and user experience of the app, and communicated the technical possibilities of Apple's app development framework. Based on that, we built the first version of Pic Prune natively in SwiftUI and took it to the App Store in two weeks.

Pic Prune uses on-device AI models, provided by Apple. No photo leaves the user's device and no account is needed to use it. They wanted Pic Prune to be able to predict which photo a user would be most likely to select from a group, and have that prediction model learn over time. This was the most technically challenging problem.

In later versions of Pic Prune, we implemented a one-time payment for pro features like unlimited photo comparison, statistics, and a "Fast Review" mode, that turns photo comparison into a feed-like experience, similar to TikTok or Instagram Reels.

Pic Prune comparing two near-identical photos side by side, with a note naming the one it picked as the best shot in the group and the reasons why. — Comparison view: the prediction model's pick, and the reasons it gave.

Fast Review in Pic Prune: one photo from a group with the app's pick marked, the rest of the group beneath it, and a button to keep the chosen one and send the others to the trash. — Fast Review: one card per group, with the next photos prefetched.

We shipped it in eight languages, including a right-to-left layout for Arabic, which increased the potential user base by hundreds of millions worldwide.

After launch, the team wanted to find alternative ways to drive traffic to the App Store, so we built a companion website with a blog, and developed a content marketing strategy to boost App Store downloads.

The picprune.com home page: the headline "The app that keeps your photo library under control", an App Store button, and a phone showing two photos being compared. — picprune.com: a support site and blog, built alongside the app.

The outcome

Pic Prune is live on the App Store with a 4.9 rating, in eight languages, under 11 MB to download, and at version 2.1 with the paid tier in place. It collects no data at all, which is the biggest achievement, given that we are in a world where apps are collecting all kinds of information about users. In addition to the app itself, the website works as a lead generation funnel to boost downloads.

Pic Prune went from an idea in a founder's mind to an app on the App Store in two weeks. Our team prototypes rapidly because the people who design it are the people who build it. If you have an idea for a mobile or web application, reach out.