An attribution model is a set of rules that defines how credit for a conversion (purchase, sign-up, etc.) is distributed among the multiple marketing channels and touchpoints a user interacted with along the way.
For example, if a user follows the path Google Search -> Social ad -> Email newsletter -> Purchase, the attribution question is which channel deserves credit for that purchase.
There are six major models. Last-click (100% credit to the final touchpoint), first-click (100% to the first touchpoint), linear (equal credit to all touchpoints), time-decay (more credit to touchpoints closer to conversion), position-based (40% each to first and last, 20% split evenly among the middle), and data-driven (machine learning calculates optimal distribution).
Shortened URLs serve as a critical data source for attribution analysis. By using different shortened URLs (with distinct UTM parameters) for each channel - social media, email, ads - you can accurately track which channel's link was clicked and ultimately led to a conversion. Decide how granular the channel split should be before you start, or links for the same campaign end up scattered across several short URLs and can no longer be added up.
Google Analytics 4 uses data-driven attribution by default. This model leverages machine learning to calculate each touchpoint's actual contribution based on historical data. As of August 2026, however, GA4 itself offers only data-driven attribution and last-click variants (paid and organic last click, and Google paid channels last click); the first-click, linear, time-decay and position-based models were retired in November 2023. If conversion volume is too low for data-driven attribution to feel reliable, compare it with a last-click view, and export the data into a spreadsheet or BI tool when you want to test other distributions.