Shaping algorithmic visibility through unplanned communication: recommendation systems, customer segmentation, andethical drift in digital marketplaces

This study aims to explore the interplay between unplanned, user-generated communication and recommendation systems in determining product visibility, customer segmentation, and inequality within digital marketplaces. Drawing on prior work in recommendation systems, business intelligence, and algorithmic decision-making, the study articulates a framework outlining how early user engagement acts as a strategic input into systems that allocate visibility. Analysis leverages product-level interaction data and counterfactual simulations to explore whether engagement at an early stage is associated with higher visibility at later stages, whether recommendation systems may amplify these effects, and whether exposure driven by algorithms is associated with outcomes of segmentation and inequality. This is an explicitly associational and simulation-based rather than a causal analysis, seeking to identify patterns consistent with amplification, concentration and visibility inequality in digital marketplaces. The findings further provide evidence that early user-generated communication is positively correlated with future product visibility and that recommendation systems, particularly those based on popularity signals, appear to strengthen this relationship. In addition, popularity has self-reinforcing dynamics over time and ranking mechanisms are associated with more concentrated exposure around a relatively small number of products. Findings also suggest that different user segments face different sets of products; and concerning inequality, they suggest that even similar products may experience unequal visibility. Overall, the study highlights how unplanned communication monitored by analytic systems and processed with algorithms helps shape digital marketplaces as algorithmically mediated visibility systems.