MARCO framework improves ad conversion prediction by decomposing click intent

Researchers have developed MARCO, a framework designed to improve ad conversion prediction by decomposing user clicks based on intent. Unlike traditional models that treat all clicks equally, MARCO categorizes clicks by intent, leading to more accurate predictions for high-intent clicks and a reduction in over-prediction for low-intent ones. This approach has demonstrated a 2.80% increase in conversions per click and a 0.98% improvement in overall topline metrics when deployed. AI

IMPACT Enhances ad conversion prediction accuracy by differentiating user intent, potentially leading to more effective advertising campaigns.

RANK_REASON This is a research paper detailing a new framework and its performance metrics. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• click-through rate
• MARCO
• post-click conversion rate
• reinforcement learning

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →