Paper Number

ICIS2025-2172

Paper Type

Short

Abstract

Online content platforms face a critical design choice between slide mode (single-item display) and feed mode (multi-item display) for their recommendation systems. We develop a stylized model of a two-sided platform to analyze how this choice interacts with the recommendation algorithm's strength and creator incentives. Our analysis reveals a fundamental trade-off: slide mode has a higher per-view monetization ability, while feed mode compensates for weaker algorithms by outsourcing the final content matching to the user. Our key finding is the existence of a performance threshold based on algorithmic capability. Feed mode generates superior revenue when the algorithm is weak; however, as the algorithm strengthens, the matching advantage of the feed mode diminishes, and the slide mode's superior monetization makes it the dominant choice. Practically, our work provides actionable insights, demonstrating that platforms must align their interface design with their algorithmic maturity to optimize performance.

Comments

15-Interaction

Share

COinS
 
Dec 14th, 12:00 AM

Slide or Feed: Content Recommendation with Creators' Incentives

Online content platforms face a critical design choice between slide mode (single-item display) and feed mode (multi-item display) for their recommendation systems. We develop a stylized model of a two-sided platform to analyze how this choice interacts with the recommendation algorithm's strength and creator incentives. Our analysis reveals a fundamental trade-off: slide mode has a higher per-view monetization ability, while feed mode compensates for weaker algorithms by outsourcing the final content matching to the user. Our key finding is the existence of a performance threshold based on algorithmic capability. Feed mode generates superior revenue when the algorithm is weak; however, as the algorithm strengthens, the matching advantage of the feed mode diminishes, and the slide mode's superior monetization makes it the dominant choice. Practically, our work provides actionable insights, demonstrating that platforms must align their interface design with their algorithmic maturity to optimize performance.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.