An attribution model is the rule for assigning credit for a conversion to the individual touchpoints with the customer. It decides whether credit for a sale goes to the first ad, the last, or all of them along the way, and so it directly affects which channels you judge effective and where you shift budget. The choice of model can change the assessment of the same campaign by tens of percent, which makes it a strategic decision, not a technical detail. Simple models are rule-based (e.g. last click), more advanced ones are data-based (algorithmic, regression-based attribution, Markov chains). (Compare: RBA, Markov chain)