Project by: Bishal Sharma Roy, Cherrie Yeung,
🎯 Objective
To implement a data-driven personalization framework for PepsiCo's diverse product portfolio by leveraging consumer persona modeling. The project aims to demonstrate how AI can be used to tailor product recommendations based on demographic, psychographic, and taste-related attributes.
🧪 Methodology
- Product Catalog Analysis:Â Structured PepsiCo product data was categorized by attributes including flavor profile, product type, and nutritional positioning.
- Persona Design:Â A controlled subset of 10 consumer personas was developed with defined demographics and preferences.
- Recommendation Mapping:Â Each persona was linked to relevant products using keyword-based semantic matching from product descriptions.
- Generalization Model:Â This methodology is scalable for the full product range using identical classification, segmentation, and AI-copy generation frameworks.
👥 Persona-Product Mapping (Representative Segment)
Persona Table
📈 Impact
- Enables hyper-personalized digital campaigns across global markets.
- Improves product discoverability by aligning with consumer intent.
- Lays the foundation for predictive targeting through behavioral data.
🔠Future Scope
- Automate persona-product pairing using NLP classifiers and recommendation algorithms. Integrate with CRM systems to deliver dynamic product experiences at scale. Incorporate A/B testing frameworks for performance validation of AI-generated copy.