Title: Attentional Funnel Mapping: Deep Neural Attribution for Stage-Aware Marketing Abstract: We introduce a deep sequential architecture that learns to map raw user behavior trails to latent purchase funnel stages using global attention. Our funnel-aware attention mechanisms assign interpretable, advertiser-specific importance scores to activities—enabling both accurate conversion prediction and automatic behavior tagging. The resulting neural attribution system outperforms existing models in offline AUC (+3.6% lift) and aligns closely with human stage annotations. In production A/B tests, funnel-stage targeting drove 3–6% incremental conversion lift, proving that jointly modeling journeys and funnel states enhances both prediction and personalization in digital advertising.