In what ways could Substack improve its discovery system to help readers find new writers and publications they might enjoy?

Introduction

Improving Substack's discovery system to help readers find new writers and publications they might enjoy is a critical challenge for the platform's growth and user engagement. As we dive into this product improvement case, we'll explore various aspects of the current system, identify pain points, and propose innovative solutions to enhance the discovery experience for Substack users.

Step 1

Clarifying Questions (5 mins)

• Looking at Substack's position in the content creation market, I'm curious about the current user engagement metrics. Could you share some insights on the average time spent on the platform and the frequency of new writer discoveries per user?

Why it matters: This information will help us understand the baseline for user engagement and the effectiveness of the current discovery system. Expected answer: Users spend an average of 15 minutes per session, discovering 1-2 new writers per month. Impact on approach: Low discovery rates would indicate a need for more aggressive recommendation algorithms and UI changes.

• Considering the diverse range of content on Substack, I'm wondering about the content categorization system. How granular are the current content categories, and how do users typically navigate between them?

Why it matters: The depth and breadth of content categories directly impact the discovery process. Expected answer: Substack has broad categories with some subcategories, but navigation between them is limited. Impact on approach: A more refined categorization system might be necessary to improve discovery accuracy.

• Given the platform's focus on supporting independent writers, I'm interested in understanding the balance between promoting established writers and surfacing new, emerging voices. What's the current approach to this balance in the discovery system?

Why it matters: This balance is crucial for both reader satisfaction and platform growth. Expected answer: The system currently favors established writers with larger followings. Impact on approach: We might need to adjust the algorithm to give more visibility to promising new writers.

• Considering the potential for personalization, I'm curious about the data Substack currently collects and utilizes for user preferences. What types of user data are being leveraged for content recommendations?

Why it matters: The depth and quality of user data significantly impact the effectiveness of personalized recommendations. Expected answer: Substack uses basic data like reading history and newsletter subscriptions. Impact on approach: We might need to implement more sophisticated data collection and analysis methods to improve personalization.