UF study: Social media posts can predict which product features drive sales
Researchers at UF’s Warrington College of Business developed an AI framework that analyzed more than 3 million Instagram posts alongside sales data from a kitchenware company to identify content likely to trigger purchasing behavior. The team created a metric called purchase-evoking frequency to measure how often specific product features appear in those posts, finding a significant positive link to actual sales. The study suggests businesses can use this approach to inform product design, inventory planning, and supply chain decisions rather than relying on broad engagement figures like likes or shares.
Sources: UF News