Avatar AI: Case Study
How Canan AI automated Avatar AI's repetitive workflows, and unlocked the one core roadblock in their business model to skyrocket top line revenue by 2000%.
New York, NY, US
2022
Consumer Digital Products
$10+ billion (2023)
20+
Challenge
The original business model had a structural roadblock in the number of avatars that could offer monthly subscriptions due to time spent communicating with users, manually creating engaging content, and overseeing day to day operations. The time intensive workflows and lack of a clearly scalable model made it challenging for firms to drive revenue significantly, oftentimes with revenue plateauing and falling flat after a certain number of subscriptions was reached. Our task was to fundamentally uproot the core business model to allow firms to continually add AI avatars that didn't require hours of human time spend to create and maintain to allow scalability.
Results
Post-redesign, the number of AI avatars was increased by 20x with a 95% reduction in time spent on repetitive workflows to maintain those avatars. This meant a potential 20x in revenues. User engagement significantly increased with over 53K monthly impressions across social platforms. The streamlined workflows and modern interface resulted in a 80% increase in new subscriptions within the first 3 months of the launch.
95%
Decreased time on repetitive workflows
80%
Increase in monthly subscriptions
20x
Increase in potential revenues
Process
Research & Analysis: We explored SaaS product options such as Zapier & Gumloop that offered AI automation services. We dove into the potential integrations these services offered that would be applicable to our clientele.
Information Architecture: Based on the research findings, we created an information architecture model of the company's fundamental workflows, with each step being represented by a node (similar to how Large Language Models (LLMs) are structured), to accurately visualize which nodes could be fully automated by AI.
Wireframing & Prototyping: We designed low-fidelity wireframes to visualize the new layout and navigation of these workflows, iteratively refining them based on user feedback. Afterward, we built a high-fidelity, interactive prototype to test the design.
Usability Testing: We conducted usability tests with a diverse group of users to validate the design and identify areas for improvement. Users data was further split into cohorts to analyze cohort retention data separately given how they responded to automation via A/B split tests. Based on the feedback, we made necessary adjustments to the automated workflows.
Results & Future Integrations: We developed a series of workflows that integrate AI automations to reduce time spent on repetitive tasks by up to 95%. We also created Standard Operating Procedures (SOPs) and templates of how to recreate the workflows and to maintain design consistency in future updates.
“ When I started Avatar AI, I had no idea if the concept was even profitably viable. But, when I discovered the power of AI automations, I instantly knew I would be able to create something special. Ultimately, it lead to starting Canan AI which now delivers those same AI automations that I used myself to firms around the world which has proven critical in reducing labor costs and skyrocketing revenue. ”
Shreyas Goswami
CEO, Founder | Avatar AI
Conclusion
Whether you are in the business of AI avatars, or insurance brokerages, AI automations have the potential to skyrocket your topline revenue while also reducing your labors costs. Through automating repetitive workflows, insurance agents will be able to significantly boost their productivity by focusing on high leverage tasks, enabling firms to streamline their workforce and reduce labor costs without compromising on service quality or client satisfaction.