This document discusses attribution modeling and how to move beyond simplistic single-click attribution models. It notes that the consumer purchasing journey can span many interactions over time. The current measurement model treats attribution as a single event, but the real customer path is very complex. Rather than seeking a perfect attribution model, the goal should be actionable insights that reduce uncertainty and enable better decision making. The document outlines a framework for attribution that includes discovery, testing, evaluation and iteration to generate hypotheses, test attribution models, and share learnings. It also discusses how TagMan customer and behavioral data can be used to power attribution modeling and visualization.