Google Ads and Google Analytics Attribution Models Guide

This article describes Google Ads and Google Analytics attribution models in detail. After reading you will be able to choose the right attribution model for your advertising strategy and determine which channels or ads result in sales, are worth investing your budget in, and which channels can be dropped altogether.

An attribution model is a principle of distributing conversions, transactions, and their monetary value between their sources, it’s especially important for Google Shopping Ads Management.

At this point, we can track a huge amount of data after a user clicks on a Google ad. The data is recorded through the gclid tag that’s added automatically to the ad link, for example:

gclid=Cj0KCQjwnP-ZBhDiARIsAH3FSRcr0LrJXWrRtROUQt9sqMH4Qc0zDN1f_cMaK69FTsIzIM69QxpgCSgaAvXpEALw_wcB

The gclid tag is generated automatically after the ad is clicked on. It sends data to the system, such as the search query, date of the click, data about the ad that was clicked, etc. The information received through the gclid tag will be stored in your browser’s cookie.

Before making a conversion (call, form submission, purchase, and so on), the user can go a long way before that. The user may go from one ad to another, looking for a good deal. Eventually, when the user makes a conversion, the established tracking tag processes the cookie and sends it to the advertising account.

Let’s have a look at the attribution options available in your Google Ads account when setting up conversions.

Last Click: Assigns all the value for the conversion to the last ad and the corresponding keyword that was clicked.

First Click: Assigns all the value for the conversion to the first ad and the corresponding keyword that was clicked.

Linear: Distributes the conversion value per conversion evenly among all ad interactions along the way.

Time Decay: Assigns more value to advertising actions that occurred closer in time to the conversion. The click value doubles every week.

Position-based: Assigns 40% of the value to the first and the last click, with the remaining 20% distributed among the remaining clicks.

Data-driven: Allocates value per conversion based on the data already collected. It’s different from other models and uses collected data to calculate the actual value of each interaction on the conversion path. Data-driven attribution is used by default for most actions-conversions.

Let’s break down the “Data-driven” attribution model, because it is the most universal. At the moment, this attribution is available for most advertising accounts, including new ones.

You can check the availability of this attribution by clicking on the “Switch to DDA” tab in your Google Ads settings:

1. Provides information on which keywords, ads, ad groups, and campaigns play the most important role in achieving your business goals. Distributes value at a specific point in the customer journey.
2. Optimizes bids based on account performance data. 
3. This attribution model is the most versatile and fits most business models.

Watch this video to learn more about using data-driven attribution for Google Search:

In Google Ads this attribution model follows two principles:

• The system develops models of conversion probabilities for all transition paths 

It uses data on users who have or have not made a conversion. The system analyzes this data and evaluates how likely a user will make a conversion at a specific point on the path. An attribution algorithm based on the data evaluates each Google ad shown. The result is a model that shows how likely a user is to make a conversion at each waypoint for a given sequence.

• The system assigns values to all methods of interaction with users

The data-driven attribution algorithm takes into account the sequence in which each mode of user interaction is present and assigns a different coefficient for all transition stages. For example, one model is created for the case where a display of a media ad precedes a Google Search ad, and if they are in reverse order, a different model is used.

These principles also allow you to connect a “Data-driven” attribution model for new accounts without accumulated data.

Here’s a great example from Google: birthday attribution

Source: ThinkWithGoogle

Who made the birthday party a success?

Source: ThinkWithGoogle

When was the birthday party planned?

Source: ThinkWithGoogle

Time lag.

Source: ThinkWithGoogle

Note

Data-driven attribution is not the ultimate cure-all. If you think the last click is the most valuable for your business, then use “Last click,” and so on.

Have a look at the attribution options available in the Google Analytics account when analyzing data.

Last Interaction: The latter source is assigned the conversion and all its value.

Last Non-Direct Click: The last source is assigned the conversion and all its value, but if it was a direct link to the site, the value is assigned to the source before it.

Last Google Ads Click: The last source from Google Ads is assigned a conversion and all its value. 

First Interaction: The first source is assigned the conversion and all its value.

Linear: Distribution of conversions among all sources equally. 

Time Decay: Assigns more value to actions that occurred closer in time to the conversion. The click value doubles every week.

Position Based: Assigns 40% of the value to the first and last click, and the remaining 20% is distributed among the other clicks on the path.

Note

You may have noticed that in the attribution models listed above, there is no “Data-driven” model. At the moment, you cannot select this attribution model in your Google Analytics account. It is only available for analysis in the paid version of GA-360 analytics. So if you have a “Data-driven” attribution model enabled at the conversion level in your Google Ads account, then the comparative analysis of Google Ads account data and Google Analytics data may differ, and you should keep this in mind.

What are the differences and common mistakes when comparing Google Ads and Google Analytics attribution models?

The main difference is in the data displayed in the accounts. A Google Ads account only receives data related to advertising traffic (single-channel system). A Google Analytics account receives data from different channels (multi-channel system).

In Google Analytics, the default attribution model is the Last Non-Direct Click. Therefore, there is a “Model Comparison Tool” in your Google Analytics account, which allows you to compare different attribution models.

If you don’t see any attribution models in your Google Analytics account, you can enable them yourself and create custom models (by moving the slider to the right):

The most common mistake is an inaccurate evaluation of the advertising channel due to the initially chosen attribution model. We have already described above how to compare attribution models.

For example, when analyzing ads with an attribution model based on the last click, you can find out that some campaigns don’t bring conversions (at the bottom of the funnel) compared to the other ad campaigns, and make the wrong decision —  you disable these campaigns. But they could have played a very important role at the top of the funnel. In this case, all the other campaigns would stop getting the necessary targeted traffic, and their performance might worsen.

Ideally, you need to test different attribution models and see what’s right for your business. Testing attribution models usually takes a lot of time, and you don’t always have it. That’s why we recommend using the “data-driven” attribution model.

If you already have conversion data in your Google Ads account, there is an opportunity to apply tools to pick up an attribution model. Go to the “Attribution” tab in settings:

The first is an attribution modeling tool. It allows you to compare two different attribution models (and compare other attribution models with “Data-driven”):

The second tool is the main paths before the conversion. Here you can see how users interact with ads before converting:

Choosing the correct attribution model is an important step for properly analyzing your advertising channels. It allows you to interpret the results correctly and make effective strategic decisions. 

Check out other valuable articles and guides in our e-commerce PPC library.

Our agency specializes in providing top-notch PPC lead generation management services to help businesses increase their online visibility and drive more conversions. If you still have questions about attribution models, or which model to choose, don’t hesitate to contact our team.