When setting up conversions, Google offers you different attribution models, in order to understand which click has the most credit for a conversion. 

That’s why Google Ads offers you 6 different attribution models.

1. Last Click Attribution Model 

This is the default attribution model google uses when setting a campaign for the first time, without touching the attribution model section. Like its name is, whatever your campaign goal is (lead, conversion), the result will be attributed to the last click on your ad. 

Let’s say you run 2 campaigns. One is in search, and the other is in display. If you get a conversion in display, you might think that the client saw your ad and converted right away. But does this really happen in reality? How much is taken into account regarding the research done before?  See why the Last Click model is too partial for one campaign/ad/keyword, offering too little information on the performance and the implications of your other campaigns too. 

2. First Click Model 

Concentrates on what channel, keyword, or ad actually drove traffic to your site first and ended in a conversion. 

First Click model also has its limitations. If, for instance, you run a search campaign and a user clicks on your keyword but doesn’t convert. Instead, when they see a remarketing ad following them across the internet, that is when they convert– In this scenario, the conversion is attributed to that first click. 

So again, you are missing the potential of your remarketing campaign 

3. Linear Model 

Every interaction the user takes before converting is counted equally.

A linear model is beneficial if you run long-term campaigns, starting with general keywords and moving to remarketing campaigns. It is also preferred for products and services that need a little research before acquisition and that move from awareness to conversion actions. 

4. Time Decay Model 

Gives more credit to clicks closer to a conversion. 

If it took 14 touches for a client to convert, the last few ones will get all of the traffic. For the time decay model, you need complex and long-term campaigns. This is because the closer you get to a conversion, the actions before the conversion are counted in a decreasing manner. Credit is distributed using a 7-day half-life.

5. Position-Based Model 

Gives 40% credit to the first and 40% to the last click; the remaining 20% is given between the other clicks from the client journey.

So the first and the last click have a heavy lift, whereas the middle keywords, along the consumer journey, help you get a picture of what combination of awareness keywords lay between the first that drove engagement and the last keyword, which in fact, closed the sale.

6. Data-Driven Model 

Is only available to accounts with enough data. 

It distributes credit for the conversion, taking into account machine learning and past data for that conversion. That’s why this model is based on account performance.