Introduction
Youth Impact is a youth-led, evidence-based organization working across multiple countries in Southern Africa, with a mission to scale proven programs in education and health to benefit children and youth at large. With a footprint in Botswana, Namibia, South Africa, the Philippines, and a major presence in India through partnerships like J-PAL and the Karnataka government, the organization has reached over 225,000 youth to date. At its core, Youth Impact is committed to two principles: scaling impact and using data to learn and iterate.

One of their flagship interventions is the “Teaching at the Right Level” (TaRL) program, a globally recognized, evidence-based approach that targets foundational learning gaps in children by grouping and teaching them at their actual learning levels. This method has demonstrated significant improvements in learning outcomes, and Youth Impact has adapted and scaled it across several countries.

However, a central challenge has been the need for reliable and scalable data collection to monitor student progress and inform teaching. To address this, Youth Impact turned to Glific—a WhatsApp-based open-source chatbot platform—to streamline image collection from teachers. This case study outlines how Glific transformed their data operations and what lies ahead.

At Youth Impact, paper tools were a deliberate and foundational element in implementing the “Teaching at the Right Level” program. Teachers used these tools daily to assess students’ current learning levels, group them by ability, and tailor lessons accordingly. These tools also captured student-level data to inform decisions and track program outcomes. While paper tools were low-tech and scalable in the field, they presented a major challenge in scalability when it came to digitizing the data. Teachers submitted images of the completed tools via WhatsApp, which were then manually entered into data systems—a process that became increasingly unsustainable as the program scaled to over 100,000 children annually.

---

The Challenge: Data at Scale and Speed
Two critical components of the TaRL approach are:

1. Ability-Based Grouping: Teachers must first assess each student individually to determine their learning level.
   
2. Data-Driven Teaching: Teachers ask a “question of the day” after each lesson and use students’ performance to decide whether to proceed or revisit the content.
   

These assessments are documented through structured paper tools, which guide daily teaching decisions and serve as a key data source for Youth Impact’s research and learning teams.

As the program scaled to over 100,000 students annually (doubling or tripling year over year), managing this volume of student-level data became increasingly complex. Teachers submitted these forms by capturing images and sending them via WhatsApp to regional officers, who then handled manual data entry into SurveyCTO. This process, while effective in small settings, quickly became a bottleneck.

---

The Glific Solution: Centralized, Scalable, and Structured
To solve this, Youth Impact implemented Glific to:

• Centralize the image-sharing with live dashboards in Gsheets
• Manage incoming files (automatically named and sorted) in the Gdrive 

Here’s how the process works:

1. Teachers send the keyword “TaRL” to initiate the chatbot.
   
2. Based on their phone number, the chatbot recognizes or updates their school and region.
   
3. Teachers are guided to upload images of paper tools, selecting if the data is for baseline or endline assessments.
   
4. These images are validated for file type and quality, and links are logged into a central Google Sheet.
   
5. A structured dashboard allows the Youth Impact team to track submissions, assign mentors, flag issues, and prepare for manual data entry into SurveyCTO.
   

---

Interactive Demo and Features
Emilie from Youth Impact demonstrated the chatbot’s functionality, showing how it:

• Filters school selection dynamically through pre-loaded region and alphabet filters.
  
• Allows multiple image uploads in a single flow.
  
• Tracks each submission with complete metadata: contact, location, type of data, and file links.
  
• Sends gentle nudges for incomplete or low-quality submissions.
  
• Publishes directly to a live dashboard for downstream data entry and quality control.
  

---

Laying the Groundwork for Machine Reading
The eventual goal is to use machine learning via Google Vision to automatically read and digitize these paper tools—eliminating the need for manual entry altogether. To achieve this, centralization of image submission via Glific is the first and necessary step.

Over the next 12–24 months, the team will:

• Feed Glific-submitted images into Google Cloud Vision.
  
• Compare machine-read results with manually entered data.
  
• Transition to fully automated data digitization once accuracy thresholds are met.
  

---

Lessons Learned

1. Teacher Adoption: Teachers rated the chatbot’s ease of use at 9.6/10—unusually high for a data tool—showing strong acceptance.
   
2. Structured Piloting: A phased pilot approach allowed the team to iterate on flows, fix bugs, and scale with confidence.
   
3. User Behavior Insights: Common queries like “How do I group students based on level?” are now being built into an interactive helpdesk within the chatbot.
   
4. Efficient Scaling: Standardizing image sharing through Glific is a critical precondition for scalable, high-frequency data operations.
   
5. From Paper to Predictive Analytics: The project showcases a unique hybrid model of low-tech, field-friendly tools (paper forms) backed by high-tech, scalable backend systems (Glific + Google Cloud).
   

Future Scope:

• Soon to come: An interactive help desk for FAQs
• In development: Integration with Google Cloud to machine read all their paper tools for maximum scalability.

---

Conclusion
Youth Impact’s journey with Glific exemplifies how chatbots can bridge the last-mile data gap in education, making large-scale monitoring both feasible and efficient. Their implementation highlights that when designed thoughtfully, technology can enhance—not replace—the work of educators on the ground, while providing organizations with the tools they need to scale with impact and insight.