Google Analytics Embraces Data-Driven Attribution

Introduction

Google Analytics, the popular web analytics service that tracks and reports website traffic, is making some changes to its attribution models in October 2023. Attribution models are the rules that determine how credit for sales and conversions is assigned to touchpoints in conversion paths. 

Google Analytics currently offers several attribution models, such as last-click, first-click, linear, time decay, position-based, and data-driven. However, Google has announced that it will be removing four of these models: first-click, linear, time decay, and position-based. These are all rules-based models that assign value to each touchpoint based on predefined rules. For example, the first-click model gives all the credit to the first touchpoint, while the position-based model gives 40% of the credit to the first and last touchpoints and 20% to the middle touchpoints.

Why is Google removing these models?

According to a Google spokesperson, the reason is that these models "don’t provide the flexibility needed to adapt to evolving consumer journeys." Google also claims that these models have "increasingly low adoption rates, with fewer than 3% of conversions in Google Ads using these models." Google suggests that switching to the data-driven attribution model typically results in a 6% increase in conversions for advertisers.

What is the data-driven attribution model?

It is a model that uses advanced artificial intelligence (AI) to understand the impact each touchpoint has on a conversion. Unlike the rules-based models, the data-driven model does not rely on fixed rules or assumptions. Instead, it analyzes data from your account and learns from the performance of your ads across different devices, browsers, locations, and other factors. It then assigns credit to each touchpoint based on how much it contributed to the likelihood of a conversion.

What are the benefits of using the data-driven attribution model?

According to Google, the data-driven attribution model can help you:

• Optimize your bids and budgets more effectively by focusing on the touchpoints that drive conversions.
• Gain deeper insights into how your customers interact with your ads across different channels and devices.
• Create more relevant and personalized ads based on the customer journey stages.
• Improve your return on ad spend (ROAS) by allocating your resources more efficiently.

What are the challenges of using the data-driven attribution model?

While the data-driven attribution model sounds promising, it also has some drawbacks that you should be aware of. For instance:

• The data-driven attribution model is not transparent. You cannot see how the model assigns credit to each touchpoint or what factors it considers. This can make it difficult to explain or justify your results to stakeholders or clients.
• The data-driven attribution model is not consistent. The model may change over time as it learns from new data and adapts to changing consumer behavior. This can make it hard to compare your performance across different periods or campaigns.
• The data-driven attribution model is not customizable. You cannot adjust or modify the model to suit your specific needs or preferences. You have to trust that the model is doing its best to reflect your reality.

What are some alternatives to using the data-driven attribution model?

If you are not satisfied with the data-driven attribution model or if you are not eligible for it, you still have some options to choose from. For example:

• You can use the last-click or Google paid channels last-click models. These are simple models that give all the credit to the last touchpoint or the last Google paid touchpoint respectively. These models are easy to understand and implement, but they may undervalue other touchpoints that influence conversions.
• You can use calculated metrics. This is a new feature that Google Analytics is introducing along with the removal of the four rules-based models. Calculated metrics allow you to combine standard or custom metrics using mathematical formulas. You can use calculated metrics to create your own custom attribution models based on your own rules or logic.
• You can use third-party tools or platforms. There are many other tools or platforms that offer different types of attribution models or solutions. Some examples are Adobe Analytics, Facebook Attribution, Mixpanel, Segment, Wicked Reports, and more. You can use these tools or platforms to complement or supplement your Google Analytics data and insights.

How to prepare for the changes in Google Analytics attribution models?

If you are currently using one of the four rules-based models that Google Analytics is removing, you should take some steps to prepare for the changes. For example:

• You should audit your reports and identify how the removal of these models will affect your data and analysis. You should also compare your results with the data-driven attribution model or other models to see how they differ.
• You should test different models and scenarios to find the best fit for your goals and objectives. You should also evaluate the impact of changing models on your bidding and budgeting strategies.
• You should communicate the changes to your team, clients, or stakeholders. You should explain the reasons and benefits of switching to the data-driven attribution model or other models. You should also provide guidance and training on how to use and interpret the new models.

Conclusion

Google is replacing these models with the data-driven attribution model, which uses advanced AI to understand the impact each touchpoint has on conversion. The data-driven attribution model can help you optimize your bids and budgets, gain deeper insights, create more relevant ads, and improve your ROAS. However, the data-driven attribution model also has some challenges, such as a lack of transparency, consistency, and customization. You can use calculated metrics or third-party tools or platforms as alternatives to the data-driven attribution model. You should audit your reports, test different models, and communicate the changes to prepare for the industry updates in Google Analytics attribution models.