Retail industry executives gathered to discuss artificial intelligence’s evolving role in their sector. The panel, moderated by Chris Walton of Omni Talk, featured Pierre-Yves Calloc’h from Pernod Ricard, Vipin Gupta of Starbucks India, and Sarah Kunst of Cleo Capital.
AI: Hype vs. Reality
The conversation kicked off with a debate on AI’s current hype level. Gupta argued AI is “highly overhyped,” citing the gap between potential use cases and actual implementation. He emphasized the need for organizations to build proper foundations in data, infrastructure, and change management.
Kunst offered a nuanced view, suggesting that existing AI applications in areas like supply chain are “under-hyped,” while generative AI may be “overhyped.” She cautioned against inflated expectations, noting that AI still struggles with many real-world tasks.
Calloc’h shared success stories from Pernod Ricard, including an AI-powered application that streamlines product catalog management for distributors. He stressed the importance of finding the right use cases that deliver measurable business impact.
Computer Vision and Retail Analytics
Gupta highlighted Starbucks India’s use of computer vision to track customer behavior in stores, measuring metrics like queue times and order fulfillment speed. While showing promise, he noted ROI challenges when scaling such technologies across all locations.
Calloc’h mentioned Pernod Ricard’s application of computer vision for safety compliance in production facilities, demonstrating AI’s potential beyond direct revenue generation.
Content Generation and Brand Identity
The panel explored AI’s role in content creation, with Kunst warning that AI-generated content can sometimes feel “flat” and potentially clash with established brand identities. Calloc’h countered that AI tools are enhancing, not replacing, human creativity in marketing, allowing for faster iteration and more options.
Media Mix Optimization
Calloc’h shared how Pernod Ricard leverages AI for marketing mix modeling, enabling more precise budget allocation across channels. This data-driven approach has led to measurable improvements in marketing effectiveness across multiple markets.
Personalization at Scale
Gupta detailed Starbucks India’s efforts to deliver personalized experiences through AI, including tailored app messages and product recommendations. He emphasized the importance of foundational elements like data structures and infrastructure to enable these capabilities.
Build vs. Buy Dilemma
The panel disagreed on whether companies should build or buy AI capabilities. Kunst advocated for buying solutions in most cases, citing the complexity and data requirements of AI development. Gupta countered that for companies like Starbucks, building in-house can be more effective and cost-efficient. Calloc’h highlighted the potential of open-source solutions as a middle ground.
Back Office Applications
Calloc’h emphasized AI’s potential in streamlining back-office operations, particularly in data normalization and invoice processing. He advised a measured approach to implementation, focusing on one feature at a time to ensure proper adoption and change management.
As AI continues to evolve, retail and CPG companies face both opportunities and challenges in harnessing its potential. Success will likely depend on careful strategy, robust infrastructure, and a willingness to experiment and adapt.