Expanding Lingvist
Lingvist is an adaptive language-learning platform using AI
Designing Lingvist's Custom Decks, pivoting from educator tools to personalized language learning, driving engagement.
Timeline
June 2018 -> June 2019
Role
Product Designer
Platforms
Scale
Startup
Overview
Lingvist was originally designed as a consumer-focused language learning platform, but unexpected demand from educators in Arizona led us to explore a new market. At the same time, existing users were eager for more content beyond the ~5,000 words in our standard course. These two challenges—supporting educators and expanding content—culminated in the creation of Classroom and Custom Decks.
User interviews
To design an effective tool for educators, I conducted qualitative research with language teachers. This included:
- Interviews to understand their workflows, challenges, and goals.
- Journey mapping to visualise their planning, teaching, and assessment processes.
- Pain point analysis to identify areas where technology could improve their experience.
Key findings:
- Teachers found lesson planning and creating custom materials time-consuming.
- Many still relied on traditional textbooks but were open to technology.
- Managing different knowledge levels within a class was a major challenge.
- They needed engaging, adaptable content that could be shared with students.
This research directly influenced the design of Custom Decks, ensuring it met real classroom needs. The insights also helped refine the product’s messaging and onboarding.
Early work
Initially, our focus was on helping educators create custom vocabulary courses for their students. Through user interviews and workflow mapping, we identified key pain points, such as the time-consuming nature of lesson planning and the challenge of keeping students engaged.
To address these, I sketched early concepts for a teacher-focused tool, allowing them to quickly generate and share courses. These sketches explored:
- Creating custom course content by
- Entering single words.
- Importing passages of text for analysis.
- Student progress tracking to help teachers assess learning outcomes.
Pivoting to Consumers
As we developed this idea, we realised that scaling in the edu-tech market was more challenging than expected. Adoption barriers, such as school bureaucracy and teacher workload, made it difficult to gain traction. Meanwhile, our analytics and user feedback highlighted a strong demand from individual learners who had already completed our general course and wanted more content.
This led us to pivot from an educator-focused tool to a consumer-driven experience. Instead of teachers creating courses for students, we empowered any user to create and customise their own learning decks—giving them greater control over their language learning journey.
This shift allowed us to:
- Tap into a larger, more engaged user base.
- Improve retention by offering unlimited learning opportunities.
- Leverage organic growth loops as users shared their decks with others.
By repurposing the backend from our teacher-focused version, we could quickly launch and validate this new approach with an MVP on Android.
Solution
With Custom Decks, users could:
- Create a personalised course by scanning or inputting text.
- Adapt it by removing words or modifying sentences.
- Learn from AI-generated sentences tailored to their content.
To refine the experience, we ran guerrilla testing sessions in our London office and remote surveys. A key insight was that the term "Publish" confused users, leading them to believe their decks would be made public. We adjusted the copy, improving conversion rates.
Outcome
Custom Decks launched with strong engagement:
- 48% of users said they would be very disappointed if it were discontinued.
- 81% felt the example sentences were "just right."
- The end-of-flow completion rate improved from 14% to 64%.
- Android retention rates increased.
We also identified a potential growth loop: users could share decks with friends, bringing new learners to Lingvist.
Conclusion
Working on both Classroom and Custom Decks provided deep insights into how learning happens at scale. We launched Classroom at a teachers’ conference in New Orleans and later introduced Custom Decks to consumers, where it had a more immediate impact.
This project was a defining moment in my career—working with AI and linguistics, and shaping a key feature that launched on Product Hunt. It was a challenging but rewarding journey that redefined Lingvist’s approach to content creation.