THE DOMINO EFFECT IN CONSUMER DECISION-MAKING : PREDICTING SEQUENTIAL PURCHASE BEHAVIOR USING AI AND ANALYTICS

Abstract

The domino effect in consumer decision-making refers to a sequential purchase behavior where an initial purchase triggers a cascade of related purchases, shaping retail dynamics and marketing strategies. This study explores this phenomenon among North Indian consumers, leveraging a dataset of 133 responses collected via Google Forms. The dataset encompasses demographic details (age, gender, income), purchase channels (online, physical stores), motivations (necessity, convenience, marketing influence), and purchased products (e.g., smartphones, laptops, cameras). By employing Artificial Intelligence (AI) techniques—Sequential Pattern Mining, Recurrent Neural Networks (RNNs), and Decision Trees—the research aims to identify, predict, and explain sequential purchase patterns, such as a smartphone purchase leading to accessories like cases or earbuds. The central research question is: How can AI and analytics predict and explain sequential purchase behavior within the domino effect framework? The objectives include identifying prevalent purchase sequences, predicting subsequent purchases, and evaluating the influence of demographic factors and motivations. The methodology integrates Sequential Pattern Mining to detect frequent purchase sequences, RNNs to forecast follow-up purchases, and Decision Trees to assess the impact of variables like income and purchase intent. Data analysis revealed actionable insights into consumer behavior in a localized context, with implications for both businesses and academic research. Methodology and Key Findings The study utilized a three-pronged AI approach. Sequential Pattern Mining identified common purchase sequences, such as "Smartphone → Case" with 33.8% support, indicating a strong tendency for accessory purchases following primary product acquisitions. The RNN model, trained on sequential purchase data, achieved an accuracy of 83.5% in predicting subsequent purchases, demonstrating robust predictive capability. Decision Tree analysis underscored the significance of income (particularly 3-10 LPA) and necessity as primary drivers of sequential purchases, with e- commerce recommendations amplifying the effect. Key findings highlight that necessity-driven purchases, coupled with targeted online recommendations, fuel the domino effect among North Indian consumers. Middle-income consumers emerged as a dominant segment, frequently extending smartphone purchases into accessory ecosystems. These results validate the efficacy of AI in decoding complex consumer behavior patterns, offering a granular understanding of how initial purchases catalyze subsequent buying decisions. Implications and Conclusion For businesses, the findings suggest practical applications in optimizing recommendation systems and designing product bundling strategies to enhance customer engagement and sales. By anticipating sequential purchase patterns, retailers can strategically position complementary products, improving cross-selling opportunities in digital marketplaces. For researchers, this study lays the groundwork for future investigations, such as scaling the analysis with larger datasets or exploring advanced AI techniques like reinforcement learning.