Guide to Attribution Models in Google Analytics 4 (GA4)

Customers’ path to the conversion is complicated and nonlinear at times: users come and go and return to the site again using various sources such as organic searches, referral links, or an ad targeted to a user in one of the social networks. Here come attribution models in GA4 to determine which source made the greatest contribution before making a purchase or completing another important action on the site.

In the beginning, to understand how attribution works, you will have to figure out how to choose and set the proper attribution model, what is the lookback window and why there are different scopes. Let’s delve into our expert guide to Google Analytics 4 attribution models.

Key Takeaways:

• An attribution model determines how the credit is assigned to touchpoints that lead to conversions. Understanding Google Analytics attribution is a must for any data-driven marketer. It can help you make profound decisions about where to allocate your marketing and advertising budget.
• Dimension scopes are an important part of the correct interpretation of data.
• The attribution model set by default is a Data-driven model. It can be changed in the admin panel. But keep in mind that this change affects only the event-scoped dimension.
• Choosing the right attribution model requires thoughtful analysis and experimenting. Unless you notice big differences in the data of the Model comparison or Conversion path report, it is best to keep the default attribution model in your GA4 account.
• Remember that attribution modeling is not a one-time decision. Regularly review and update your attribution model based on changes in your business environment, marketing strategies, and customer behavior. It’s a dynamic process that requires ongoing analysis and adjustment.

Table of Contents:

1. What is a GA4 Attribution Model?
2. Google Analytics 4 Attribution Models
3. What Do Scopes Have to Do With Attribution?
4. How to Change the Attribution Model in GA4 and How Does This Change Affect the Data?
5. Where to Find Attribution Reports in GA4?
6. What Else Should You Know to Deal With the Attribution?
7. GA4 Attribution Models in Practice
8. What is the Best GA4 Attribution Model?
9. Conclusion

What is a GA4 Attribution Model?

A GA4 attribution model is a set of rules or a data-driven algorithm that assigns credits to touchpoints along the user’s complex journey of conversion. A touchpoint is an interaction with the site throughout a marketing channel, and the combination of touchpoints that leads a customer to a key action is called an attribution (or conversion) path.

Different ways of processing data (different attribution models) result in different outcomes.

Why is it important?

First of all, as a business owner, you should see how your marketing channels contribute to conversions and sales. Understanding which channels (paid search, social media, email, or organic search) are effective in driving customer actions can help you to allocate the budget more efficiently.

Secondly, utilizing attribution models will also help you to optimize campaigns, improve ROI, and make data-driven decisions.

And finally, effective attribution models enable businesses to understand the customer journey better, leading to more targeted and personalized marketing strategies.

Now, let’s proceed with the different attribution models available in GA4.

Google Analytics 4 Attribution Models

Currently, in GA4, there are three attribution models available:

1. Data-driven attribution
2. Paid and organic last click
3. Google paid channels last click

Note: you can find the resources mentioning more than 3 models, but they were deprecated in November 2023 and are not available in GA4 anymore.

1.1 Data-driven attribution model (DDA)

The model set in the GA4 property by default is a ‘data-driven attribution’ model. It uses the machine learning algorithms and historical data of your specific account. DDA model processes both converting and non-converting paths and determines how different types of touchpoints impact the outcome. The model also takes into account different factors such as device type, number of ad interactions, user activity on the site, type of ad’s creative, and much more.

The model’s algorithm uses path data to understand how the presence and timing of the touchpoint affect the probability of key action taken. And then, based on the calculated rates, it assigns credits to touchpoints.

DDA can attribute the conversion to multiple sources at once. For example, depending on whether a user sees an ad in a search or not, the conversion probability can increase by 50%.

1.2 Paid and organic last click model

Probably the most common and easy-to-understand model is ‘Paid and organic last click’, also known as ‘Last non-direct click’. It ignores the direct traffic and attributes the conversion and its value to the latest touchpoint before conversion.

For example, with the next touchpoints path.

The conversion will be 100% attributed Organic Search.

1.3 Google paid channels last click model

‘Google paid channels last click’ model, similar to ‘Paid and organic last click’, attributes key actions to the last touchpoint, but if there is a Google Ads in the user’s path it will receive gives credits only to Google paid channels. If there is no Google Ads touchpoint, the model works just as  ‘Paid and organic last click’ one.

Using the last click model, especially if you advertise on other platforms, you may get the wrong impression that your paid campaigns have a lower conversion rate than Google Ads.

Note: All attribution models exclude direct visits from receiving attribution credit unless the path to the key event consists entirely of direct visits.

Pros & cons of various allocation models

Each attribution model has advantages and disadvantages that must be considered when analyzing conversions. Here is a summary table:

What Do Scopes Have to Do With Attribution?

Regarding attribution, we cannot mention the traffic source dimensions in GA4: channel, source, medium, campaign, etc. The traffic-source dimensions provide information about the traffic coming to the site, where it is coming from, what method is used, and which marketing effort is responsible for its acquisition.

Scope is a characteristic that describes the extent to which a GA metric or dimension can be applied. 

GA4 organizes traffic information into three groups: user, session, and event-scoped. With scope in GA4, Google can apply attribution credit at different levels of granularity. Starts with the smallest scope of the conversion event itself, then looks at the session during which the conversion has occurred, then zooms out to look at the user’s whole journey and how they originally reached the site. Understanding scopes is fundamental to building, interacting with, and interpreting reports in GA4.

The scope can be distinguished one from another by the prefix: there’s now a Session Source dimension, a First User Source dimension, and a standalone Source dimension.

2.1 User-scoped dimensions

User-scoped dimensions are associated with the very first user’s source and don’t change during the user’s lifetime (unlike session and event scopes).  These dimensions always include the prefix “First user”.

2.2 Session-scoped dimensions

Session-scoped dimensions are set when a new or returning user starts a new session.

These dimensions always include the prefix “Session”. New values are assigned to session-scoped dimensions each time users open the site or app.

2.3 Event-scoped dimensions

Event-scoped dimensions help you attribute credit for a key event that a user triggered. Event-scoped dimensions don’t include a prefix, as in Source or Medium. The source and medium for non-key events are “(not set)”.

Keep in mind that each dimension should be combined with the metrics of applicable scope. For example event-scoped dimensions can be used with Key events and Total revenue, but using it with such metrics as Session or Users will not work as intended.

2.4 Comparison

Below is the summary table with the scopes comparison.

Example: A user visited the site 4 times using different sources and made a purchase during the latest session.

If you look at the dimensions of different scopes, you will see the next values:

How to Change the Attribution Model in GA4 and How Does This Change Affect the Data?

The model set by default in GA4 is Data-driven attribution. If you prefer a different model, you can change it in the Admin panel.

To select attribution settings:

1. Sign in to Google Analytics
2. Open Admin panel
3. Under Property settings, expand the Data display section
4. Click Attribution settings
5. Select the Reporting attribution model

Changing the reporting attribution model applies to historical and future data.

Please note changing the reporting attribution model is reflected in all key event reports and explorations that use event-scoped traffic dimensions, for example, Source, Medium, Campaign, and Default channel group. 

User- and session-scoped traffic dimensions, such as Session source or First user medium, are unaffected by changes to the reporting attribution model. These dimensions follow the ‘Paid and organic last click’ model, regardless of what you select.

Where to Find Attribution Reports in GA4?

Different standard reports in GA4 use various attribution models depending on the scopes of dimensions included in these reports. Here are the standard GA4 reports:

4.1 ‘User acquisition’ report

Where to find:

• ‘Reports’ tab -> ‘Life cycle’ section -> ‘Acquisition’
• ‘Reports’ tab -> ‘Business objectives’ section -> ‘Generate leads’

Why use this report? With the User acquisition report you can get insights into how new users find your website or app for the first time.

4.2 User acquisition cohorts

Where to find:

• ‘Reports’ tab -> ‘Life cycle’ section -> ‘Acquisition’
• ‘Reports’ tab -> ‘Business objectives’ section -> ‘Generate leads’

Why use this report? As well as a User acquisition report, you will find out how new users find your website or app for the first time.

4.3 Traffic acquisition’ report

Where to find:

• ‘Reports’ tab -> ‘Life cycle’ section -> ‘Acquisition’
• ‘Reports’ tab -> ‘Business objectives’ section -> ‘Generate leads’

Why use this report? Easily understand where your website and app visitors are coming from. It specifically shows where new and returning users come from.

4.4 ‘Attribution models’ report (formerly called ‘Model comparison’)

Where to find:

• ‘Advertising’ tab -> ‘Attribution’ section

Why use this report? Compare how different attribution models impact the valuation of your marketing channels.

4.5 ‘Attribution paths’ report (formerly called ‘Conversion paths’ report’)

Where to find:

• ‘Advertising’ tab -> ‘Attribution’ section

Why use this report? Understand customers’ paths to triggering key events and how different attribution models distribute credit on those paths.

You can also create customized and exploration reports specified for your needs.

What Else Should You Know to Deal With the Attribution?

Except for all the mentioned above, make sure you have key events, formerly called conversion, and transactions (or both) correctly configured and collecting the data in GA4.

Another thing to consider is the Lookback window. It determines how far back in time the attribution model will look for a touchpoint to attribute it to conversion, as users can trigger the key events days or weeks after interacting with your ad.  The conversion window applies to all attribution models and all conversion types. Changes to the conversion window apply going forward and will be reflected in all reports within your Analytics property.

GA4 Attribution Models in Practice

Now that we have a theoretical basis for each concept, we will bring it all together and review some examples to see how credit for the conversion would be attributed under each model. 

Note: Assuming the random distribution across each touch for the data-driven model for demonstration purposes.

So let’s review a complex user journey: The user visited the site 3 times, but during the last session, he received the registration confirmation email and, navigated to the site via an email link, then made a purchase.

Example 1. Attribution model set in the property: Paid and organic channels last click. Here is what you will see in the report when comparing the dimensions of different scopes.

As you can see in this example, since the user used two sources to open the site during one session, event-scoped source conversion is attributed to the latest one, while the session source didn’t change because GA4 does not start a new session when the user changes the source.

Example 2. Attribution model set in the property: Google paid channels last click. The same path will show different results when the model is changed.

As you can see, the event-scoped channel is changed because the model gave credit to the latest paid channel according to the description.

In the table below, you can find examples of the common user paths and the expected output values of attribution models.

As you can see, in some cases, the values are very different for different models. It mostly happens when a user changes the source during the session, for example, opens an email link or enters the site using another source before the session ends in 30 minutes.

These cases can be found in real data. However, most businesses will barely face the situation when the majority of sessions contain several sources.

Need help configuring Google Analytics 4?

• Book a consultation today
• Read our success stories

What is the Best GA4 Attribution Model?

This is one of the most frequent questions we receive from clients: “What is the right attribution model for our business?”. As always with this type of question, there is no one-size fits all answer. The choice depends on business specifics: their media mix, industry, etc. We recommend following two rules:

1. Different questions require different models

Understanding your business goals and considering the complexity of customer journeys is a starting point of choosing the model. Sounds complicated, but when you read a few cases below, the idea will become more clear.

Use-case №1. Understand what marketing campaigns are best at converting users into customers if the time from site entry to purchase decision is short.

In this case, the latest session source matters the most, so you will look at the session-based dimension (‘Session source/medium’), which is the last click attribution model. Or you can check the data from a data-driven model perspective, as it also tends to give more credit to the touch points at the end of the conversion path, given the nature of its computation.

Use-case №2. Understand which marketing efforts are most effective at acquiring new customers when the customer journey is longer, and it takes some time for a customer to make a decision and return to the site for purchase.

Here, we recommend analyzing the traffic sources using user-scoped dimensions (First user source/medium’), which will give a picture of the first touch of users.

Use-case №3. Understand how different marketing tactics are working together.

If you have enough historical conversion data and touchpoint interactions to provide reliable results, you can leverage the data-driven attribution model or even go a step further and build overlap analysis:

2. Keep it simple and consistent

Sometimes, businesses delve deep into data collection and cleaning, trying to get the perfect data set, and they end up spending more time collecting the data than analyzing.

But in most cases, monitoring changes is more important than exact numbers. Pick the attribution model and use it consistently over time to observe the dynamics.

One important note: compare apples-to-apples, email campaigns among themselves, etc. In other words, apply the attribution model to comparable segments of data.

Conclusion

GA4 provides different attribution models. Understanding attribution modeling allows businesses to analyze and evaluate the effectiveness of their marketing efforts. Use Google Analytics 4 to make smarter and more informed decisions.

If you are looking for experts who can guide you through GA4, contact VIDEN, the Google Analytics 4 consultants.