Coach AI from 0-1

Overview

The goal was to design and ship an AI-powered chat bot for the FantasyPros website—a core company priority for the year. Initially intended as a beta for internal testing, the project quickly evolved into a full-fledged product thanks to an iterative approach and cross-functional collaboration. By leveraging OpenAI's capabilities, we launched a cutting-edge tool that not only enhanced user engagement but also became a successful driver of our premium tier.

Problem Statement

FantasyPros needed to explore how AI could enhance its product offerings. The challenges were:

• New Technology: Limited internal expertise with OpenAI and its applications.
• 0-to-1 Product Development: Designing a functional MVP from scratch.
• Iterative Approach: Releasing quickly while maintaining quality to gather actionable user feedback.
• Monetization: Moving quickly from a free to paying product.

Goals

• Explore and validate the potential of AI-powered tools for FantasyPros.
• Build a functional MVP that aligns with user needs and internal quality standards.
• Test and iterate based on feedback to ensure usability and value.
• Integrate the chat bot into the product suite and drive engagement and conversions.

Approach

Research & Discovery:

• Collaborated with the Data and AI team to understand OpenAI’s capabilities and identify the most impactful use cases.
• Conducted initial user interviews to pinpoint potential scenarios where a chat bot would add value.
• Worked with Data and AI team to build out the initial version of the Chatbot on ChatGPT to test functionality.

Development & Iteration:

• After validating functionality, worked with designers and engineers to create the front-end, on-platform experience, ensuring seamless integration with the existing platform.
• Designed and implemented the chat bot’s functionality using OpenAI’s GPT, fine-tuning its responses to match user needs.
• Released the product iteratively, starting with a beta version for internal testing to gather feedback.

Launch & Scaling:

• We'd only planned to release a beta version internally. After early success, we were able to release an internal version, a beta version with registered users, a premium version, and a freemium version for logged out users, build a full screen experience, and release it on the apps. This far exceeded out expectations.
• Expanded the scope to include logged users and also found a way to fully roll out a premium version of the product.
• We've continued to gauge feedback to refine bot, improving accuracy, usability, and feature set.

Results

• Engagement Boost: The chat bot significantly increased user interaction on the site, driving engagement metrics higher.
• Recruitment and Activation: After finding a way to create a freemium version of the product, we saw the tool become a big driver of registrations and activation.
• Reversed Negative Trends: Purchases for the associated plan showed a measurable uptick, reversing prior declines.
• Innovation Delivered: Demonstrated FantasyPros’ ability to leverage emerging technologies like OpenAI, positioning the company as a forward-thinking leader in the space.

Key Learnings

• Rapid Experimentation: Working with new technologies requires agility and a willingness to experiment.
• User Feedback is Crucial: Iterative development allowed us to validate assumptions and improve product quality in real-time.
• Cross-Functional Collaboration: Close coordination with data, AI, design, and engineering teams was critical to success.
• AI in Practice: Building with OpenAI required not just technical expertise but also a strong focus on use-case alignment.

‍