Case Studies

Planteria

Snapshot

Not everyone like's to read to much nowadays (thanks TikTok). See a comic strip?

Comic Strip

Snapshot

A plant care app that learns from its community

2023

Product Designer, Buses & Cabs

Tools: Sketch · After Effects · Illustrator

Interviews

80+

Survey Responses

76.25 to 82.27

Jump in SUS

The Problem

There are hundreds of photo apps, and almost none of them help you take a better photo. Every suggestion arrives afterwards, in edit mode; which is exactly the wrong moment. The critical time is while you're holding the camera up.

Our starting concept came from a shared interest in music, photography, and film: recreate famous photos and movie scenes based on where you're standing. That grew into something wider: templates, pose guidance, and real-time feedback.

Research

10 Interviews

With people across a range of photography skill, friends and family included. What devices, when they shoot, what frustrates them.

People overwhelmingly use phones, shoot most when travelling or somewhere new, and want something nice enough to share — but not everyone's a professional and the effort is real.

Qualtrics Survey

80+ responses, asking what apps they use, what their goals are, which features they'd want, and to rank those features. We selected nine essential features from it.

Ideas (everything, good or bad), Product Goals (merged into real-world actions), Users (scenarios)

80+ responses, asking what apps they use, what their goals are, which features they'd want, and to rank those features. We selected nine essential features from it.

Competitive Analysis

Across Instagram, Snapchat, TikTok, VSCO, AiryCam, and Ulike confirmed the gap: nobody was doing real-time guidance during capture.

The decision

WE CUT THE COMMUNITY FEED

For a photo app this is close to heresy; social is the obvious growth engine, and every competitor we studied leaned on it. We chose content creation over social networking and deliberately left the feed out. The product does one thing: help you take a better picture. Adding a feed would have made it a worse version of six apps that already exist.

WE KEPT TEMPLATES CONTRARY TO SURVEY RESULTS

Only 38% ranked photo/video templates as high priority, under our own bar. We kept the feature anyway, because it was the differentiator and because a survey measures what people can imagine, not what they'd use once it existed. The concept tests backed that up: pose recommendations and spot suggestions were what people responded to most.

Process

Crazy 8s

Eight sketches each in eight minutes, then presenting to find common themes. Camera flow, search, and map integration surfaced immediately.

Then wireframes, then concept testing with hand-drawn sketches deliberately kept rough so people understood it was a work in progress and felt free to criticise it.

Medium Fidelity

Moderated usability testing over Zoom and Figma, thinking aloud. SUS: 76.25. Above the 68 average, not good enough. We fixed four things: specific search instructions like "Search Barton Springs", a progress indicator, clearer real-time feedback, and a browsing function in maps.

What was shipped

High Fidelity Designs

Building upon our mid-fidelity screens, we elevated the design to a high-fidelity level. We developed a design system, incorporating a color palette and iconography for a cohesive visual identity. High-fidelity screens aimed to enhance the app's visual appeal while maintaining user-friendliness.

Unmoderated Usability Testing

With high-fidelity screens in place, we conducted unmoderated usability testing using UserTesting.com. This phase involved 10 participants who completed specific tasks within the app. The feedback gathered helped fine-tune the app, addressing issues like instruction clarity and user understanding of features. Achieving an SUS score of 82.27 indicated improved usability

The result

Interviews

80+

Survey Responses

76.25 to 82.27

Jump in SUS

Feedback was strong enough that some participants asked when they could actually use it. Error, action, and success colours were praised for clarity, and people specifically liked that the step messaging felt human. Icons around editing and saving still confused people.

NOTE FROM FUTURE SAMARTH

We tested with 10 and got a good SUS score, and I'd treat that with more suspicion now. A camera app lives or dies in real conditions; bad light, moving subjects, one hand full. Every test we ran was someone calmly holding a phone in a quiet room.

Always up for a conversation about UX, design systems or AI-native tooling

Download Resume