< Home
Stamp Cutout Effect cover — a grid of photos made with the tool

Overview

Stamp Cutout Effect — Making a viral photo trend easier for anyone to create

A free web app that helps people recreate a stamp-cutout photo effect without downloading another app or learning professional design software.

Role

Product Design
Design Engineering

Timeline

2026

Tools

Figma
Claude Code
Codex

Skills

Product Strategy
0→1 Product Design
Interaction Design
AI Code Prototyping

Impact

4K

Unique visitors

In roughly 3 weeks after launch.

48%

Traffic from South Korea

Share of total traffic, driven by one reshare.

40+

Daily visitors

Still arriving after active promotion slowed.

Vercel Analytics: 4,184 visitors and 6,149 page views, a spike peaking at 1,127 visitors, and South Korea at 48% of traffic
Vercel Analytics across the first three weeks

The Problem

The effect was simple. Creating it wasn’t.

I noticed an Instagram trend where people placed a stamp-shaped cutout of a photo above the original image. Existing apps could recreate it, but they often required downloads, multiple editing steps, and limited control. Professional design tools offered more flexibility, but came with a much larger learning curve.

The trend as it circulated, and the multi-step workaround people followed to recreate it in an existing app

Opportunity

How might I make this effect quick and easy to create with fewer steps—without requiring an account or app download?

Process

01 · From AI-generated starting point to a designed product

I used AI-assisted coding to generate the first working version based on the features I wanted. The initial output gave me a functional starting point, but the interface felt generic and needed stronger visual hierarchy, clearer interactions, and more intentional product decisions. I redesigned the UI and refined the experience through repeated iteration.

The AI-generated first version: a two-panel interface with a heavy control sidebar
Before — the AI-generated first version: too much UI for a simple task, and the canvas wasn’t the focus.
The redesigned interface: a slim canvas and stamp sidebar beside a large working canvas
After — the redesign: controls reduced to canvas and stamp, with the composition given the space.

02 · Five deliberate design decisions

Rather than adding more features, I focused on a small set of decisions that kept the experience simple and predictable.

Controlled resizing

Stamp size is limited to 340×240–450×318 with a fixed 17:12 ratio to preserve the visual balance of the effect.

Stage-based controls

Users can zoom the photo before stamping and resize the stamp before finalizing, reducing accidental edits.

Automatic photo fitting

Uploaded photos automatically fit to the canvas width, giving users a usable starting composition immediately.

One-cutout limitation

The experience supports one stamp at a time because the trend is centered around highlighting one favorite part of a photo.

Responsive interactions

Desktop uses mouse and resize handles, while mobile uses pinch and drag for interactions that feel natural on each device.

Final Prototype

Launch

Launching gave me the feedback I couldn’t get from designing alone.

I first shared Stamp Cutout on Reddit and X. Reddit gave me early product feedback, including the resize request. Later, I shared the tool with a South Korean design influencer who was posting about the same trend. They tried it, liked it, and reshared it with their audience.

The Reddit resize request, my launch post on X, and the Korean influencer's reshare
The Reddit resize request, my launch post on X, and the reshare that carried it further

User Feedback

Posts from Korean users sharing stamp cutouts they made with the tool
What people made with it, in their own posts

Reflection

What I learned

AI is a starting point, not the final designer.

AI helped me move quickly from idea to prototype, but the product improved through deliberate interaction decisions, visual constraints, and repeated iteration.

Distribution is part of product building.

Shipping the tool wasn’t enough. Sharing it where the trend already existed is what connected it with users.

Real feedback is more valuable than guessing.

The resize feature came directly from a user request, and the product became better because I shipped early enough to hear it.