Shopee stands as the foremost e-commerce platform across Southeast Asia and Latin America, where like its counterparts, it utilizes deep learning techniques to enhance its ads recommendation systems. Different types of ads recommendation scenarios emerged to satisfy the needs of different users. Among these, Trigger-Induced Recommendation (TIR) is a relatively new field that has recently gained attention for targeted optimizations from both the industry and the research field. In this work, we pinpointed a problem specific to TIR scenarios in the industry, named Trigger Relevancy and Diversity Inefficiency (TRDI). Illustrating with Shopee Product Detail Page You May Also Like (PDP YMAL) recommendation section, where billions of item impressions occur daily, we introduce a systematic approach for the industry community to examine the extent of TRDI in their TIR scenarios. Then we propose a novel approach: Dual-Phase Synergistic Attention (DPSA) method, to tackle the TRDI problem. We proved the effectiveness of DPSA by integrating it with Shopee PDP YMAL recommendation section's Ads CTR ranking model. After thorough online A/B testing and holdout testing, DPSA has become the new baseline at Shopee PDP YMAL PaidAds CTR ranking across all regions of Shopee markets, significantly enhancing key metrics for user experience, seller profits, and platform gains.