Swadhaar

Research and Iterations

This project adopted a diagnostic approach focusing on usability to map out different iterations for the Swadhaar chatbot. This learning tool tackles subjects of financial literacy such as using UPI, fraud prevention and awareness, and accessing online banking services. This approach ensured that the solution was not stopped at enabling access to financial literacy, but rather focused on behaviour change to push towards adoption for financial inclusion. The focus of the digital learning platform was to

• To understand the context, the people, their motivations, digital prowess and identify potential entry points for a digital tool
• To understand which types of users to design for
• To gauge what are the most popular and requested financial queries people have and the barriers to accessing such information and products
• To understand the role that community trainers and champions play and discover how trust is established

Methodology

We adopted a diagnostic research framework, combining in-depth interviews to map user friction points with field staff intercepts to understand the frontline realities of digital onboarding and trust-building. By synthesising these qualitative insights with quantitative data, we conducted usability testing on alternate chatbot flows to refine tone, modality, and sequencing to ensure the platform was not only accessible but also emotionally resonant for long-term engagement.

To smoothen the learning process, we tested three iterations to understand user behaviour and preferences of our target user group:

Gamification: We tested whether digital safety champions could be promoted using badges and progress bars. While these tools offered a sense of achievement for some, they were often associated with correctness rather than progress, this highlighted a need to design for celebrated milestones.

Menu selection: We explored shifting from push-based flow of information to menu-based interactions, to help test if users preferred to pull information on-demand. These options had a mix of videos, questions, achievement badges and informational text.

Chatbot: To drive higher engagement, we experimented with an AI chatbot trained on Swadhaar content, allowing users to ask questions in their own natural language. All users could interact with the content in Hindi, English as well Hinglish but the content was delivered in Hindi. Users were also able to access videos, audios/voiceovers and text.

Opportunities for Scale

As the project moves toward a 5K client roll-out, several design directions have emerged to ensure the chatbot acts enables the goal of financial adoption amongst its user base.  Since many users typically scroll rather than search on WhatsApp, we are redesigning the chatbot’s identity to be relevant and familiar to users by building an identity of a Sathi  (partner) for trust building. Additionally, the quality of a client’s initial interaction is defined by their experience with the trainer. As a result, we are developing bitesized training toolkits which are 2-3 minutes long and serve as practical guides that equip trainers to handle real-world logistical challenges and ensure a consistent onboarding experience.

The project thus works through an iterative experiment approach where qualitative insights helped us in reframing digital financial literacy as a social and emotional journey, rather than just a functional transfer of information, and a commitment to scalability.