struggling to accurately implement multi-touch attribution modeling for long conversion cycles, specifically data reconciliation issues

we're trying to refine our marketing analytics setup, specifically around how we credit touchpoints and optimize spend. the main technical block is reconciling disparate data sources for our multi-touch attribution modeling; conversion paths often span months, leading to significant data drift when stitching together sessions from Google Analytics, ad impressions, and CRM events for a single user journey. what are the best technical approaches or tools for accurately aligning these long-tail, cross-platform user paths for precise attribution modeling without massive data reconciliation issues?