Redesign of the Search Experience on BYJU’S App
In this case study, I am going to talk about entire process of how we simplified the search experience with visual search functionality.
The BYJU'S app has been a leader in ed-tech, offering interactive video lessons, quizzes, and exercises to millions of students. As the platform grew, some features needed a closer look. As a product designer, I was tasked with redesigning the search experience that is used by over 100,000 students daily.
The Problem
Search is one of the most important features in the BYJU’S app. With a huge library of content, students often depend on search to find specific topics, lessons, or questions.
However, it’s been a long time since the search was last updated. As the app grew, the search experience became outdated and no longer met the needs of students. Students had to navigate through multiple screens, leading to distractions, and making it difficult to find the specific topics quickly. Also the search experience lacks features like auto-suggest and recent searches.
These are directly impacting student engagement and learning outcomes, resulting in higher bounce rates and lower satisfaction.
The goal was clear: Improve the search experience to make it faster, easier, and more effective for learners to find the content they needed.
My Role
I was the sole designer and handled the entire design process. I closely worked with the product manager, a senior designer, and engineers to ensure that the final output matched the design exactly.
Before making any design changes, we needed to understand how students were using the search, motivations behind each discovery type, and What are their pain points? What are the challenges they face during the search process? Gaining these insights was crucial for shaping our design strategy.
We conducted user interviews and usability tests with learners from different backgrounds who used BYJU'S as well as other e-learning platforms.
This research helped us uncover several key behaviors:
How students search: What actions they take when initiating a search, whether it's typing long questions or navigating through content categories.
What they search for: The types of content they seek, such as video lessons, practice questions, or topic-specific explanations.
When they want to search: The contexts in which search is used—during study sessions, while revising for exams, or when they need quick answers.
The biggest pain points of learners while using search:
Distractions from Irrelevant Content: Students often found themselves distracted by irrelevant results during their search.
Lack of Real-Time Help: There was no opportunity for real-time interaction or help during study sessions.
Complex Navigation: The search process required navigating through multiple screens.
Typing Difficulties: Long and complex queries, especially in math and science to find answers for the questions, were challenging to type out.
Ideation
With clear understanding of the problems, we started brainstorming exercises to capture ideas to address them. The goal was to capture as many ideas as possible and then narrow them down to the most effective solutions.
Here are the Two best ideas stood out:
Voice Search: it will allow students to speak their queries instead of typing.
Image Search: Students can take a picture of their question instead of typing it out and get answers quickly.
We decided to move forward with Image search, for a few key reasons:
Simplicity: Students could easily scan handwritten or printed questions, which was particularly useful for complex math and science problems where typing equations would be time-consuming.
Challenges with Voice search: Voice search had its benefits, but we identified several issues. Students might use complex educational terms, which could lead to inaccurate search results.
By choosing Image Search, we could offer students a faster, more intuitive way to search without adding unnecessary complexity to the app.
However, rather than implementing it as a separate feature, we chose to add it into the main search bar. Which is similar to Google’s search interface. This will allow learners to choose between type or scan their queries — without leaving the core search experience.
Here is the User flow for the New search experience
Design Process
We added a search bar to the home screen to eliminate the multiple step process for searching. The new search experience aimed to solve the key issues identified in our research by offering two main options: Image Search and Text Search.
1. Image Search
We designed Image Search to make it easier for students to ask complex questions without typing them out.
The interface was kept simple and familiar—like a basic camera app—so students could take a picture of their question and get instant answers.
Key features include:
Instructions and Cropping Tool: To guide students, we added instructions to help them capture only one question at a time. The cropping tool allows them to focus on specific parts of the image, ensuring accurate results.
Results Page: The Results Page was also designed to be straightforward, it includes a feedback option for getting real-time feedback from the users. Which can improve our solution experiences and the "Ask Another Question" button is added to keep the learning flow uninterrupted.
My Questions: In our research, we get to know that students often wanted to revisit previous searches. We added a "My Questions" section, making it easy for them to access their past queries.
2. Text Search
For Text Search, our goal was to make it faster and more efficient by incorporating features inspired by common search patterns.
Recent and Popular Search Terms: When students click on the search bar without typing anything, they see recent and popular search terms. This helps them quickly jump into relevant content
Suggested and Autocomplete Search Terms: As students start typing, they receive suggestions and autocomplete options, much like Google's search functionality. This saves them time and effort.
Categorized Search Terms: Search results are categorized into videos or questions, making it easier for students to find the specific type of content they’re looking for.
Instead of showing courses and educators, the results focus on video lessons, study materials, question, and explanations - exactly what students need in the moment
The Result Page
We also redesigned the Results Page to make it easier for students to navigate through large sets of data. Previously, videos, lessons, and questions were all mixed together in a single list, which made it difficult for students to find what they needed.
The new design introduced tabs to categorize the results:
The "Learn" tab focuses on video content, which is ideal for visual learners.
The "Find Solution" tab highlights questions and explanations, catering to students who want quick answers and solutions.
This will allow students to quickly switch between different types of content without feeling overwhelmed, giving them the flexibility to learn in the way that suits them best.
Impact
The redesigned search feature had a significant impact on user engagement and overall app performance. These improvements were tracked over a 3 month period following the rollout:
Content engagement rates increased by 27% across the platform. This metric was measured by tracking how often users interacted with video lessons, study materials, and other content after using the search feature.
Bounce rates decreased as students were able to find relevant content more quickly.
The visual search feature became particularly popular among younger students. This feature not only improved their experience but also boosted the overall satisfaction with the app.
Future Scope
Looking forward, there are several exciting possibilities for further enhancing the search experience:
"Ask a Tutor" Integration: Option within the results page that allows students to connect with a tutor in real time if they need additional help. This feature could bridge the gap between self-learning and personalized guidance, making the learning experience even more supportive.
Multilingual Support: Expanding the search feature to support multiple languages, it will make it more accessible to a wider range of learners.
We also improving the accuracy of the visual search by enhancing the image recognition engine. This will ensure that even complex questions are recognized correctly, leading to more precise and relevant results.
Conclusion
In the end, it was a good, great learning experience. There were so many discussions and rounds of iterations. Learned a lot of new stuff at every stage of the process. I would say it has given me valuable learning and leadership experience.