EatLens
App
A mobile app that scans food product barcodes to deliver instant nutritional information and personalised health recommendations. Designed for Indian consumers navigating complex food labels.
Role
Product Designer
Collaborator
Yuvraj Soni
Timeline
Jan 2025 to Mar 2025
Status
Published
Cross-functional Team
2 designers
Tools
Full visual documentation
See every screen, prototype and motion design on Behance
A lens on what you are actually eating.
EatLens started from a simple observation: most Indian consumers have no idea what the food they buy every day actually contains. Not because they do not care, but because food labels are dense, misleading, and designed for compliance rather than comprehension. The app scans a product barcode and immediately surfaces what matters: the ingredients worth knowing about, the nutritional values in plain language, and a personalised recommendation based on the user's health profile.
The product also supports multiple family member profiles so a single user can switch context during a shopping trip and get recommendations tailored to each family member's dietary needs rather than a generic average.
Point at any barcode
Verdict, under one second later
Low sugar and high fibre. A good fit for your profile.
Food labels designed for regulators, not people.
The core problem was information asymmetry. Brands are required to print nutritional information, but the format serves legal compliance, not consumer understanding. Serving sizes are arbitrary. Ingredient lists are ordered by weight but mean nothing to a non-chemist. Health claims on the front of packaging frequently contradict the ingredient list on the back.
Most users relied on front-of-pack claims exclusively
Indian consumers in our research primarily made decisions based on marketing text on the front of packaging. The actual nutritional panel was ignored by the majority of participants.
Families have conflicting dietary needs with no tool to address them
A diabetic parent, a child with a nut allergy, and a teenager tracking protein intake all shop in the same trolley. No existing app handled multi-profile shopping in a single session.
The scan-to-answer gap was the biggest friction point
Users who had tried competitor apps abandoned them because the time between scanning and getting a useful answer was too long. The design had to make that gap feel instant.
Watching people shop, not asking them about shopping.
Research focused on observational sessions at supermarkets and kirana stores. We followed participants through their actual shopping process and noted when and why they picked up products to read labels. The finding that changed the design direction most: nobody reads labels before putting something in the basket. They read them at the checkout or at home, which means the intervention has to happen before the purchase decision, not after it.
What we observed in stores
Step 01
Shelf
Product picked up
Step 02
Basket
Decision already made
Step 03
Checkout
Labels read here
Step 04
At home
Labels read here
Nobody read labels before putting something in the basket. The reading happened at the checkout or at home, after the purchase decision. So the product had to intervene before the basket, not after it.
Scan. Understand. Decide.
The interaction model is three steps. Scan the barcode. See a clear verdict on the product. Read the detail if you want it. The verdict screen was the hardest design problem: how do you communicate a complex nutritional assessment in under 2 seconds of reading time. We tested colour-coded systems, star ratings, and plain language summaries before landing on a combined approach: a dominant visual signal (safe, moderate, avoid) backed by a single plain language sentence that explains why.
Scan
Point the camera at a barcode. No menus, no search, no typing.
Understand
A clear verdict with one plain language sentence explaining why.
Decide
Expandable detail for those who want it. Everyone else moves on.
Early versions showed the full nutritional breakdown immediately. Testing showed users either ignored it entirely or got overwhelmed. The redesign led with the verdict, then surfaced expandable detail for users who wanted more. Engagement with detailed nutritional information increased significantly after this change.
The first prototype required users to set up profiles before they could scan anything. Every single test participant dropped off before completing setup. The redesign allowed anonymous scanning immediately, with profile creation prompted contextually when a recommendation would benefit from personalisation.
Verdict-first. Detail on demand. Family-aware.
The final product is a mobile-native scanning experience with a verdict-first information hierarchy. The home screen is a camera viewfinder. Scanning is immediate with no loading state visible to the user. The verdict screen delivers a clear health signal, a plain language explanation, and an expandable detail panel for users who want the full picture.
One trolley, five sets of dietary needs
→ Tracking protein intake
Barcode scan with instant verdict
Point the camera at any product. The verdict appears in under one second. The visual signal is readable from arm's length.
Multi-profile family mode
Up to five family member profiles. Switch profiles with a single tap during a shopping session. Each profile stores dietary preferences, allergies, and health goals.
Product history and comparison
Every scanned product is saved. Users can compare two products side by side or review their scanning history to understand their purchase patterns over time.
See the complete case study.
This portfolio page covers the research, design decisions, and product thinking behind EatLens. The full visual case study including motion design, bento layouts, final screens, and interaction flows is published on Behance.
View full case study on Behance
Motion design, final screens, bento layouts, and interaction flows.