Common models range from single-touch (last-click or first-click, which credit one touchpoint entirely) to multi-touch models (linear, time-decay, position-based) that split credit across the journey, to data-driven attribution, which uses statistical modeling on actual conversion data to assign credit based on real incremental contribution rather than a fixed rule.
Model choice materially changes which channels look effective: last-click attribution systematically overvalues bottom-funnel, high-intent channels (branded search, retargeting) and undervalues awareness channels (display, social, video) that influence a purchase earlier in the journey without capturing the final click.