Build a headshot micro app that actually gets used

Pick one job your app will do perfectly

Your app will die if it tries to be a Swiss Army knife. I’ve seen it happen. A photographer built a tool that promised to “enhance, retouch, color-grade, and format” a headshot. Clients opened it, felt overwhelmed, and closed it. They went back to emailing him asking, “Can you just make my teeth whiter?”

Choose one single, painful task and solve it completely. That’s your app’s job. For headshots, the top three pain points are background removal, teeth whitening, and smart cropping for LinkedIn. Pick one. Just one. If you choose teeth whitening, your entire app is about making teeth look naturally brighter—not also fixing stray hairs, smoothing skin, or adjusting suit color. You become the absolute best teeth-whitening app for professionals. A user with a single need will find you, use you in 60 seconds, and love you for it.

Design for the 90-second user, not you

You are not the user. You know what dodge and burn means. Your client, a financial advisor named Sarah, does not. She has 90 seconds between meetings. Her brain is fried. She needs to upload a headshot and get a result before her next Zoom call.

Build every screen for her panic. I tested this: the most-used version of my cropping tool had two buttons. “Crop for LinkedIn” and “Crop for Instagram.” That’s it. No manual adjustment sliders, no aspect ratio selector. Underneath, tiny text said “Advanced edit.” Only 3% of users clicked it—and those were other photographers. Your interface should feel stupidly simple to a busy professional. If you need to write an explanation for a feature, you’ve already lost.

The three must-have technical controls

Even a simple app needs guardrails. Without them, you’ll get garbage in and, inevitably, an angry email about a “ruined” photo. These three controls are non-negotiable.

• File Type & Size Limit: Block uploads over 20MB. Reject .heic or .tiff files automatically with a friendly message: “Please upload a JPG or PNG under 20MB.” This stops 80% of support tickets before they start.
• Face Detection Check: Before any processing runs, the app must confirm a face is found. If not, it stops and says, “We can’t find a face in this photo. Please try another.” This avoids charging someone for a blurry picture of their ear.
• Original Download Button: Always, always show the unedited original next to the result with a clear “Download Original” button. This eliminates user anxiety. They know they can’t permanently “lose” their file, so they experiment freely.

Why most apps fail: the background trap

This is the silent killer. You build a fantastic retouching app. A user uploads a photo taken in their dim kitchen with a cluttered bookshelf behind them. Your app perfectly whitens their teeth. But the final image still looks amateurish because of that terrible background. The user is disappointed. They blame your app. They don’t say, “The teeth look great!” They say, “It still looks unprofessional.”

The background is the foundation. If it’s bad, nothing else matters. Your single-job app must either address this directly or be brutally clear about its limits. A teeth-whitening app should have a upfront warning: “For best results, use a photo with a clean, professional background.” Better yet, partner with or recommend a background-removal service. Solve the whole problem, or clearly define the piece you’re solving.

Measure success with two simple numbers

Forget vanity metrics like total sign-ups. Track these two: Daily Active Users (DAU) and the Scream Test score. DAU is straightforward. How many unique people used your app yesterday? If you have 1000 total users but only 7 daily actives, your app is a museum piece, not a tool.

The Scream Test is more brutal. If you turned off your app tomorrow, how many people would scream? You find this by listening for unprompted feedback. Are users emailing you to say, “I used this for my team headshots!”? Are they mentioning it by name on a Zoom call? That’s the scream. It means you’ve built something essential, not just interesting. Aim for 10 screams a month. That’s a real, used micro app.

The goal isn't to build a complex platform. It's to become a tiny, indispensable part of someone's professional routine. Start with one job, make it effortless, guard against chaos, own the full problem, and listen for the screams. That’s how you build something that doesn't just exist—it gets used, every single day.