The plan to visit to the India Literacy Project (ILP) office kicked off last month when I had my first-ever conversation with their team. While I had heard about ILP’s work before, this was my chance to dive into it hands-on. The initial conversation left me buzzing with excitement; their vision for the chatbot project was not just intriguing but downright inspiring.
As the call unfolded, I found myself eagerly anticipating my upcoming two-week visit to Bangalore, where I would have the opportunity to meet the ILP team in person. One week prior to my visit, the team proposed a two-day agenda: Day one involved a visit to a school to witness career counseling sessions, allowing us to grasp onboarding and understand the users (i.e., students) better. Day two was earmarked for discussions with the ILP team about their work, challenges, and future plans.
Day 1: A School Visit to Remember
I have always enjoyed working with students face-to-face, and visiting a school with ILP was a rewarding experience. The school visit provided invaluable context, allowing me to better understand the onboarding process and the needs of the students. You can delve deeper into this heartwarming school visit by reading Sangeeta’s blog, where she recounts her first-time field experience with an NGO.
Day 2: Overcoming Bangalore Traffic and Building Connections
Day two started with unbridled enthusiasm – yes, even in the face of Bangalore’s notorious traffic. The ILP team, now familiar faces, welcomed us warmly and we started with friendly conversations. An unexpected delight was the opportunity to dust off my Telugu language skills, as it turned out that half of their team could converse in Telugu!
We hit the ground running with a presentation by the ILP team, diving deep into their work, the challenges they faced, and their future plans involving us. Some of their team members, like Padmaja, Hema joined us remotely to provide further insights.
Among the plethora of topics discussed, a few chatbot-related challenges came to the forefront:
- Low Onboarding: Many students weren’t transitioning to the chatbot after the in-person sessions at schools, where the chatbot was introduced (we also validated this information by checking the previous day’s data on Internal dashboard). ILP aimed to expand the chatbot’s presence to locations where in-person introductions weren’t possible.
- Low Engagement/Retention: Only around 5% of students revisited the chatbot after each session, and even among this group, many didn’t explore all of the chatbot’s content.
- Human Intervention within Automated Flows: Students often posed questions beyond the existing chatbot content, which counselors had to address manually, posing a significant challenge in managing individual chats every day.
After thorough discussion and a deep dive into each problem, we brainstormed several solutions and ideas:
- Addressing Low Onboarding:
- We proposed tweaking the onboarding process based on our observations from the previous day. We discussed three key aspects:
- Shortening the duration of sessions and incorporating visually engaging PowerPoint presentations tailored to capture the curiosity of younger students, leaving them eager to explore the chatbot.
- Introducing videos to illustrate how the chatbot works, enhancing students’ comprehension.
- Considering the possibility of experimenting with missed call-based onboarding, inspired by few NGOs like Reap benefit, Bharat Rohan, DG. successful NGO initiatives. Given that students often lack mobile phones during onboarding sessions, this approach could hold promise. You can read here on how Reap benefit had done onboarding through missed call and had huge registration %.
- Boosting Engagement:
- While deep diving into organization’s flows, we understood that the organization hasn’t used many Glific features like Interactive messages, Nudges in terms of no response, Triggers etc, all aimed at reducing fiction and enhancing user engagement.
- We delved into the nuances of flow and conversational design, drawing inspiration from other chatbots like Antarang, TAP, and Quest Alliance. This encompassed language usage, message length, the potential creation of a chatbot persona, and the strategic use of nudges.
3. Streamlining Human Intervention:
- We introduced the concept of Glific ticketing system, which could efficiently manage counselor-bot interactions. The ILP team showed considerable excitement about implementing it, recognizing the potential to reduce their manual workload significantly.
In a previous conversation with the ILP team, we also explored the integration of LLM and shared documentation for them to test a sample flow. During our visit, the team showcased a sample flow they had built for internal testing, and we were delighted to hear that the initial feedback was highly satisfactory.
Additionally, we presented a concise overview of the various possibilities and features of Glific, giving ILP a holistic view of the program and its potential applications.
In sum, our two-day collaboration proved remarkably fruitful for both Glific and the ILP team. We delved into multifaceted discussions, gaining invaluable insights. The ILP team expressed their intention to return with a project plan outlining the changes they’d like to implement in the chatbot and areas where they’d require our support. Their enthusiasm was palpable, and we shared their excitement in witnessing their chatbot in action once again.
I want to extend my heartfelt thank you to Team ILP (Padmaja, Rangarajan, Suriyan, Hema, Bhanu, Sabari, Karunakar,Sharan, Rajendran, Madhavan) for their warm welcome, the enlightening career counseling session, and, of course, the scrumptious Andhra meal that made the day extra special. This experience has left an indelible mark on me, and I’m thrilled about the prospect of working together to empower students and foster education through technology.