AI-Generated Flashcards
Transforming lengthy readings into bite-sized flashcards
TL;DR
AI-generated flashcards that turn long texts into concise cards, saving students' time and boosting learning efficiency
Goal
Improve app retention rate by enhancing students' learning experience
Role
Product designer (Research, UX/UI design, interactions, user testing, handoff) in collaboration with PM, engineers, QA, data analyst, UXR, content designer
CONTEXT
🎓 The business
MEET CHEGG STUDY
Chegg Study is an educational app that offers students textbook solutions, expert Q&A, and study resources to help them succeed in their coursework.
As part of the OKR we were focusing on, the team’s objective was to increase user engagement and stickiness (frequency of monthly visits) to improve our retention rate. Delivering valuable outcomes to our students was essential to achieving this goal.
🧐 The problem
INFORMATION OVERLOAD
Students often find themselves overwhelmed by large volumes of information, making it difficult to focus on what truly matters. With heavy workloads and limited time, many struggle to extract and retain key concepts from dense study materials, which impacts their overall learning efficiency.
While the TL;DR feature we released helped by summarizing long texts into concise summaries, we recognized the need for even more bite-sized learning solutions that save time, improve learning efficiency, and enhance the ability to memorize important information.
49% of TL;DR users reported that creating flashcards could enhance their productivity.
🤔 How might we help students save time and simplify complex materials to enhance their learning experience?
CORE HYPOTHESIS
01
More than reading
Reading lengthy texts is just the first step in a student's learning journey. It continues with applying, understanding, memorizing, and preparing for quizzes and exams.
02
Flexible learning needs
Students look for adaptable, on-the-go learning tools and seek customization options to tailor their study experience to their specific needs.
03
Bite-sized solution
Bite-sized learning enhances memorization by breaking down complex information into smaller, focused chunks, allowing students to retain key concepts more effectively.
SOLUTION PROPOSAL
🥁 The next step
FROM A LONG TEXT TO FLASHCARDS
_gif.gif)
At Chegg, we believe offering small learning bites will be the most effective way to re-engage users with our product.
We’ve decided to enrich the value we provide to our users by introducing AI-generated flashcards. This new feature takes lengthy study materials and transforms them into personalized flashcard decks using AI.
The feature makes memorizing key concepts much easier by breaking down long, overwhelming texts into bite-sized, focused cards. Instead of getting lost in paragraphs of information, students can quickly focus on the most important points. This format helps reduce the stress of studying by making content easier to absorb and retain.
This feature goes beyond simplifying lengthy material - it transforms how users engage with complex content, turning it into a focused, impactful study tool that enhances learning efficiency.
Hypothesis
We hypothesized that introducing the flashcards feature would drive higher engagement rates with the app, boost user retention, and enhance the overall stickiness and utility of Chegg's platform.
📋 Feature breakdown
DEFINING THE MVP
Before diving into the design, we focused on building the foundation for the new feature we envisioned. We aimed to create an MVP version, allowing us to begin testing and experimenting. We identified the following as essential components to include in our MVP:
01
Hooks
We needed two entry points from the home page and one in the TL;DR flow, along with a feature introduction to highlight the new functionality.
02
Upload material
We wanted users to have the flexibility to create their flashcards by uploading an image, document, or plain text.
03
Customization steps
We aimed to make the flashcards fully customizable, letting users set preferences like language level and detail depth to fit their learning style.
04
Gamified experience
We planned to gamify the experience by allowing users to swipe right or left to mark their knowledge, making it more fun and engaging, similar to Tinder.
05
Card interactions
We aimed to enable full interaction with cards—tap to flip, swipe to answer, navigate freely, edit, delete, or star them.
06
Deck summary
At the end of each deck, we planned to show results with options to practice wrong answers or set a reminder for later practice.
MAIN FLOW
.png)
📊 Success metrics
MEASURING SUCCESS
_gif.gif)
First and foremost, our goal is to gain quick insights into what we can offer to boost engagement and retention.
Leading indicators:
-
10% of app users will complete the entire flashcards funnel by the end of 2024.
-
Click-Through Rate (CTR) on hooks.
-
% of users uploading a document, text, or image.
-
% Success rate for flashcard creation.
-
% of users using flashcards more than once (frequency plot).
-
Average number of days visited per users.
-
Cancellation request rate.
-
North star metric - Retention rate.
DESIGN
💡 Solution overview
FEATURE SCREENS & FLOW
App homepage with 2 entry points + a walkthrough
The process of uploading an image as study material to create flashcards. Alternatively, users can upload a document (PDF, DOC, etc.), an existing image, or plain text
Flashcard customization steps that allow the user to adjust the language complexity and select the level of detail for the flashcards
Loading... Flashcards are being created
Card review and interactions: tap to flip sides, swipe to answer (or click the ✓/✗ buttons), edit content, delete a card, star a card, or skip to the previous/next card. Share the deck with friends via the '...' button.
Swiping hints for first-time users
The card design changes during the swipe action based on the response
Having a card editing option was crucial, as the content is AI-generated, and users may want to make adjustments
The results screen shows the scores and lets users review mistakes or set a reminder to practice later
Notification to remind you to practice the flashcard deck you set a reminder for.
TESTING
🕵️ User testing
TESTING THE MVP
_gif.gif)
Advancing our MVP to the testing phase received positive feedback from users who found the feature highly beneficial. They also confirmed our hypothesis: students are genuinely interested in solutions that save time and improve learning efficiency.
However, users also provided valuable feedback and some ideas for improvement:
NEXT STEPS
🤔 Final thoughts
WHAT'S NEXT?
_gif.gif)
Launch
As the feature is still in development, we plan to launch it ASAP, gather user engagement data and feedback, and then optimize the experience, fix bugs, and refine the functionality as needed.
The next phase
Simultaneously, we'll begin planning the next phase, focusing on additional enhancements and options to further enrich the user experience and maximize the feature’s impact.
