How can Thumbtack enhance its pro matching algorithm to better connect clients with the right service providers?

Introduction

Enhancing Thumbtack's pro matching algorithm to better connect clients with the right service providers is a critical challenge that directly impacts user satisfaction, retention, and the platform's overall success. I'll approach this problem by first clarifying our objectives, then analyzing user segments and pain points, before proposing and evaluating solutions. Let's dive in.

Step 1

Clarifying Questions (5 mins)

• Looking at Thumbtack's position in the market, I'm thinking about the current state of the matching algorithm. Could you share some insights on its current performance metrics, such as match accuracy or client satisfaction rates?

Why it matters: This helps us establish a baseline and identify specific areas for improvement. Expected answer: Match accuracy is around 70%, with client satisfaction at 3.5/5. Impact on approach: Lower metrics would suggest a need for fundamental algorithm changes, while higher metrics might indicate a focus on incremental improvements.

• Considering the two-sided marketplace nature of Thumbtack, I'm curious about the balance between supply and demand. Are we facing any challenges with pro availability in certain service categories or geographic areas?

Why it matters: This affects how we prioritize improvements for clients vs. pros. Expected answer: There's an oversupply of pros in popular categories but shortages in niche services. Impact on approach: An oversupply might lead us to focus on better filtering and ranking, while shortages could require improvements in pro onboarding and retention.

• Given the importance of user data in improving matching algorithms, I'm wondering about the current data collection and utilization practices. What types of user behavior data are we currently capturing, and are there any privacy constraints we need to consider?

Why it matters: This influences the potential sophistication of our algorithmic improvements. Expected answer: We collect basic user preferences and interaction data, with some limitations due to privacy regulations. Impact on approach: Rich data availability would allow for more advanced personalization, while limited data might require innovative approaches to inference and pattern recognition.

• Thinking about Thumbtack's strategic goals, I'm interested in understanding how this algorithm improvement aligns with broader company objectives. Are we primarily focused on growth, retention, or perhaps expanding into new service categories?

Why it matters: This helps us align our solution with overarching business goals. Expected answer: The current focus is on improving retention and increasing repeat usage. Impact on approach: A retention focus might lead us to prioritize improvements in match quality and user experience, while a growth focus might emphasize new user onboarding and category expansion.