Enhancing OTT Content Discovery and Engagement through the Power of Generative AI
The focus was on how voice commands, large language models (LLMs), and a CMS system unite to shape a user interface for personalised content discovery and engagement
INTRODUCTION
The brief was from a technology company based in Canada that specializes in providing managed solutions for multiscreen video services. Their goal was to revolutionize the OTT sphere by leveraging generative AI to enhance how users interact with their product, with a specific focus on improving discovery and engagement funnels.
MY ROLE
As one of the two lead designers, I played a key role in research, ideation, and creating wireframes and prototypes. Throughout this process, I collaborated closely with the product team for feasibility checks.
Additionally, I collaborated with the development team to facilitate a seamless handover and conducted quality checks to ensure alignment with the designs.
PROJECT SCOPE
A team of 3 designers(2 lead and 1 product designer), and one product manager, 2 developers. Timeline of 6 weeks.
TOOLS
Figma, Figjam
PROBLEM STATEMENT
SECONDARY RESEARCH
RESEARCH ARTICLES
We began our exploration by examining the latest articles and trends in the OTT space at the time, aiming to gather valuable insights. Some noteworthy observations included:
BENCHMARKING
We then proceeded to benchmark some of the top platforms, such as Netflix, Disney Hotstar, Amazon Prime, and others, to identify trends.
​
We conducted both direct and cross-domain benchmarking to learn from the leading platforms in the domain.
USER PERSONA
Based on the brief, we arrived at the following user persona to solve for:
IDEATION
We collaborated with the team to brainstorm ideas for integrating Generative AI into the discovery and engagement process, employing a voice-first approach to interact with the television and leveraging the existing CMS. Some of the most interesting ideas included:
MOMENTS FRAMEWORK
The next step was to chart out some of the key voice prompts that the user could give to trigger specific outputs from the system. We created a moments framework to capture interactions primarily between the user and the AI. This helped enhance the interface, providing a dynamic and engaging user experience.
SOLUTION
We started creating the solution by focusing on the shortlisted ideas. Due to the limited time and scope of the project, the client finalised two ideas from the list that focused on discovery for the purpose of the product launch. The other ideas would be taken up in subsequent sprints.
​
Incrementally, we developed mid-fidelity wireframes and progressed to high-fidelity prototypes, emphasizing the following functionality and features.​
SEAMLESS VOICE-ENABLED CHAT INTERFACE
Enabling users to discover content through the voice-enabled interface by advocating for a more conversational approach to content discovery.
INTUITIVE CONTENT DISCOVERY
Lumina transforms content discovery, making it intuitive and enjoyable by offering personalized suggestions based on user prompts, breaking away from traditional category-based exploration
TOPIC BASED CONTENT RECOMMENDATIONS
Explore content based on topics such as weather, location, seasons, festivals, and current events, nudging users to discover relevant and engaging content effortlessly
DISCOVER YOUR CINEMA PERSONA
Engaging users while facilitating content discovery, the personality quiz widget helps users identify their cinema personality and suggests content based on their personality match
DESIGNING FOR EDGE CASES
We designed for edge cases to ensure the system functions seamlessly in unusual situations, preventing any disruption to the user experience. In the context of a voice-first interaction, we considered specific scenarios like:
​
-
Cases where the AI might misunderstand the command
-
Instances where the AI may mishear the command
-
Situations where the AI cannot retrieve data for a given command
​
For this case study, I have provided a detailed illustration of the solution for the first scenario as follows:
CONCLUSION
In conclusion, we successfully addressed the given problem statement by focusing on delivering a seamless voice-based interface. Our goal was to enhance content discovery and assist users in finding the most relevant content through smart widgets and a conversational interface. The solution was showcased through a live demo at the International Broadcasting Conference (IBC) in September 2023 in Amsterdam, where it garnered interest from tech enthusiasts and entrepreneurs who interacted with the innovative solution.
​
Moving forward, we plan to implement the other ideas suggested for content discovery and engagement. The goal is to make the user experience more engaging and delightful by leveraging the capabilities of generative AI. Thank you for taking the time to read my case study!