Illustration
Rapid Prototyping
CHARGE Insurance
A bilingual concept-testing toolkit designed to evaluate motor insurance product preferences among informal sector workers in India. Working alongside BFA Global's in-house research team, my role spanned information architecture and visual communication — structuring how feature combinations were presented across four product packages, and designing the illustrated cards that made complex insurance concepts (wage loss cover, partner networks, comprehensive vs. own damage) legible to Hindi-speaking respondents with varying financial literacy. The toolkit was used as a discrete choice research instrument, with the IA influencing the quality and reliability of the data collected.
Client
BFA Global
Year
2025
Sector
Social Impact & Development
The Context
India's gig workers — cab drivers, delivery riders, auto drivers — are significantly underinsured, partly because insurance products are hard to understand. When your audience has low financial literacy and limited exposure to insurance terminology, how do you run preference research on complex product combinations?
What I Built
A pencil-first prototyping process: sketched each insurance feature by hand, photographed the sketches, and ran them through AI image generators (Gemini's Imagen outperformed Adobe Firefly and ChatGPT for this task) to produce clean line-style visuals. Iterated on prompts to standardise the visual style across features, mixed and matched elements in Photoshop, and assembled final prototype sheets in Canva with Hindi and English captions for the research team. Printed and taken to field for 25 in-depth interviews across Mumbai and Hyderabad.
Key Findings
Comprehension — Drivers compared and ranked insurance product combos without needing extended verbal explanation
Trust signal — The hand-drawn aesthetic worked in our favour: drivers engaged more openly with visuals that felt like research rather than marketing
Time cost — Five feature sets took roughly three days from first sketch to field-ready prototype, including iteration cycles and Photoshop refinement
AI accuracy — Gemini's Imagen was most precise at interpreting hand sketches and maintaining visual consistency across iterations
Back