What is an attribution model? Meaning and examples
An attribution model is the rule that decides which touchpoints get credit for a conversion. Here is what it means, the main model types, one AED sale split four ways, which models GA4 and Google Ads still offer and how to choose one.
An attribution model is the rule that decides which ads, emails and visits get credit for a sale or lead when a customer interacted with several of them first. Google describes it as "a rule, a set of rules, or a data-driven algorithm". The model never changes how many sales you made, only which channels look responsible.
What are the main types of attribution model?
| Model | How credit is split |
|---|---|
| Last click | 100% to the final touch before the conversion |
| First click | 100% to the first touch |
| Linear | Equal shares to every touch |
| Time decay | More credit to touches closer to the conversion |
| Position-based | 40% first, 40% last, 20% shared across the middle |
| Data-driven | An algorithm compares converting and non-converting paths |
| Custom | Weights you set per touchpoint |
Every model except last click and first click is a multi-touch attribution model.
Attribution model example
An illustrative AED 1,200 order follows four touches: an Instagram ad, a non-brand Google search ad, an email and a brand search ad on the day of purchase. Credit in AED:
| Model | Non-brand search | Brand search | ||
|---|---|---|---|---|
| Last click | 0 | 0 | 0 | 1,200 |
| First click | 1,200 | 0 | 0 | 0 |
| Linear | 300 | 300 | 300 | 300 |
| Position-based | 480 | 120 | 120 | 480 |
Which attribution models do GA4 and Google Ads offer?
Fewer than they used to. On April 6, 2023, Google announced that first click, linear, time decay and position-based models were going away, noting that "less than 3% of Google Ads web conversions" used them. Google Ads now offers data-driven and last click, with data-driven as the default for most conversion actions. GA4 offers data-driven, paid and organic last click, and Google paid channels last click.
How do you choose an attribution model?
Choose by the decision you are making. Leave each ad platform on its own model for automated bidding, and use one cross-channel model on de-duplicated conversions for budget decisions. Long B2B cycles need models that credit early touches. Where models disagree sharply, test with incrementality holdouts.
How Grower helps
Grower's attribution engine de-duplicates conversions from every channel and app into one unified conversion count, then credits each touchpoint with the model you pick: last click, first click, linear, time decay, position-based (40/20/40) or a custom model with your own weights.
Frequently asked questions
What is the default attribution model in Google Ads?
Data-driven attribution is the default for most conversion actions, according to Google's help center, and last click is the only other model available. Conversion actions that still used first click, linear, time decay or position-based models were switched to data-driven from September 2023, unless the advertiser chose last click instead.
Why do GA4 and Meta report different numbers for the same sale?
They use different models, windows and touchpoints. GA4 sees visits to your site and excludes direct visits from credit unless the whole path is direct. Meta credits conversions to its own ads inside click and view windows, so it can count people who saw an ad without clicking. Neither de-duplicates against other platforms.
Is data-driven attribution better than rule-based models?
It can be more accurate when there is enough conversion data, because it learns from your own paths instead of applying a fixed rule. But each platform's version only sees the touchpoints that platform can observe, and it is harder to explain. Rule-based models are transparent and comparable across tools, so many teams use both and test big disagreements.
Read the full guide to marketing attribution models, or book a demo to see Grower's attribution engine working with your own channels.