Problem Discovery (Origin of the Problem)

My exploration into this topic was sparked by Christian Thieme's insightful analysis of Udemy's recommendation system. Thieme's breakdown highlighted key areas for improvement, particularly the lack of real-time responsiveness and tailored course suggestions based on user behavior. He pointed out that while Udemy's email recommendations use a user's historical data, the real-time search experience doesn't dynamically reflect evolving interests, creating a gap in personalization that this project aims to address.

Problem Statement

How might we increase Udemy's revenue by enhancing the customer experience using AI to recommend personalized learning pathways tailored to individual user goals and skill levels?

Problem Background

Udemy's platform lacks deep AI integration to personalize the learning journey. Users often face decision fatigue due to the vast number of courses available, making it difficult for them to select the right courses. Current course recommendations rely on past user activity, which doesn't account for a learner's unique goals or professional development. Moreover, users sometimes struggle to figure out the best courses to take next based on their existing knowledge and aspirations.

While Udemy offers course recommendations based on past purchases and browsing history, these suggestions do not take into account a learner's goals, skill gaps, or overall progression. This creates a significant gap in the user experience, particularly for career-focused learners who need structured paths toward specific professional objectives.

Research Insights

User Pain Points

Interviews with 3 Udemy users and a survey of 13 respondents identified key prospective user categories, revealing:

• Decision fatigue: Users feel lost in the vast course catalog without a structured guide, leading to analysis paralysis and delayed enrollment decisions
• Skill level mismatch: Many users enroll in courses misaligned with their expertise, resulting in frustration when content is too basic or too advanced
• Career progression uncertainty: No visibility on how courses align with long-term goals, making it difficult for users to create coherent learning journeys
• Low completion rates: Users start courses but struggle to finish due to lack of motivation and clear progression milestones

Supporting Data

Research synthesis provided these key findings:

• 83.3% of users feel overwhelmed by the number of courses available, indicating a significant need for curation and guidance
• 84.6% of users desire structured learning paths to navigate their education, demonstrating strong demand for this solution
• 61.5% select courses based on career goals, yet lack personalized career-aligned guidance to ensure relevant skill development
• 53.8% of users trust AI-driven recommendations but demand transparency in how suggestions are generated
• Only 23.1% of users rely on Udemy's current recommendation system, suggesting significant room for improvement

Feedback

Preliminary user research found that users strongly desire structured learning paths with clear progression milestones. Career-focused professionals particularly valued features that could map courses to specific job requirements. Interviewees emphasized the importance of seeing a clear connection between courses and career outcomes, with one participant noting: "I want to know exactly which skills I'm developing and how they relate to jobs in my field."

Landing on the Solution

• Based on our research findings, we've developed a comprehensive solution addressing Udemy users' core pain points:
• Learning Path Generation
  
  • User Trigger: After setting career goals during onboarding, users access the "My Learning Path" dashboard to see AI-generated structured learning sequences
  • Addresses the 84.6% of users seeking clear progression toward specific learning goals
  •  Adapts paths based on user progress and feedback
• Skill Assessment Engine
  
  • User Trigger: New users take a brief skill assessment quiz, while existing users can access "Skill Evaluation" to identify knowledge gaps
  • Tackles the problem of skill level mismatch reported by 83.3% of overwhelmed users
  • Recommends appropriate starting points based on competency level
• Career Alignment System
  
  • User Trigger: User selects "Career Goals" to view how courses map to professional requirements and industry-relevant skills
  • Serves the 61.5% of users who select courses based on career advancement
  • Monitors learning effectiveness and suggests next steps
• Progress Tracking
  
  • User Trigger: Users view their "Learning Dashboard" for visualization of completed modules, skill development, and milestone achievements
  • Addresses low completion rates by providing clear progression indicators
  • Creates a comprehensive view of student advancement through defined learning milestones

‍

• Personalized Learning Pace
  
  • User Trigger: System analyzes learning patterns and suggests optimal study schedules
  • Helps improve the current 69.2% course completion rate by adapting to individual learning habits
  • Provides course completion predictions and alerts

Our research revealed these additional valuable features:

• Career-aligned skill gap visualization that clearly shows how current capabilities compare to job market requirements
• Personalized pacing options with adjustable schedules accommodating busy periods
• Alternative content suggestions when users struggle with difficult concepts, preventing abandonment
• All components integrate through a cohesive learning dashboard serving as the central hub for interaction. By combining AI-driven recommendations with user-triggered actions, the system maintains transparency while providing powerful personalization, addressing the concerns identified in our research.

‍

Key Metrics for Success

Metric

Current

Target

Course Completion Rate

69.2%

85%

User Engagement

Baseline

+40%

Revenue per User

Baseline

+30%

‍

Solution Explanation

Our solution creates a personalized AI learning ecosystem that addresses the key pain points identified in our research. The five core components work in concert to create a seamless experience from initial user assessment through continuous learning:

The Learning Path Generation and Skill Assessment Engine components establish a solid foundation by first understanding the user's current abilities and future goals. This addresses the fundamental issues of decision fatigue and skill level mismatch that 83.3% of users reported experiencing.

With this foundation in place, the Career Alignment System bridges the gap between learning and real-world application, directly addressing the career progression uncertainty highlighted by 61.5% of our research participants. This creates clear motivation for users by connecting educational content to tangible career outcomes.

Finally, the Progress Tracking and Personalized Learning Pace components tackle the completion rate challenges by providing both the visibility and flexibility users need to maintain momentum in their learning journey. By adapting to individual learning patterns, these features transform what was previously a one-size-fits-all approach into a truly personalized experience.

The integration of these components creates a virtuous cycle: as users engage more deeply with the platform, the AI gains more insights into their learning patterns, allowing for increasingly personalized recommendations. This self-reinforcing system not only improves the user experience but also drives Udemy's business goals by increasing engagement, course completions, and ultimately, revenue per user.

‍

User Flow

Future Steps

Based on our problem space research findings and solution design, we've identified these key areas for future development:

• Transparency in AI Recommendations: Adding explanation features to show users why specific courses are suggested, addressing the 53.8% of users who want more transparent AI recommendations
• Gamification Framework: Implementing achievement badges and progress rewards to boost the current 69.2% completion rate by leveraging motivation techniques
• Professional Certification Integration: Connecting learning paths to industry credentials to serve the 61.5% of career-focused users

Potential feature expansions based on identified pain points:

• Industry-Aligned Course Discovery: Using AI to categorize courses based on real-time job market demands, helping users overcome decision fatigue in the vast catalog
• Instructor Analytics Dashboard: Providing engagement metrics to instructors to improve content quality and address skill level mismatches
• Collaborative Learning Networks: Creating AI-matched study groups to enhance engagement and completion rates
• Smart Re-engagement System: Using learning pattern analysis to optimize course completion timing and reduce abandonment

High Fidelity Design

Here’s the link to the Full Design below

High Fidelity Design