Analyzing Google Optimize Experiments Data

Sharing is caring!
Google Optimize Experiment Data Analytics

Customer Data is one of the most important assets for all business domains in today’s world. Businesses are keen to understand their customer behavior, and in order to do so, they need conversion information, along with the customer reactions to specific segments.

Google Optimize is a powerful A/B testing tool that provides personalized and engaging website experiences. Its seamless integration with different Google Marketing Platform tools, like Google Analytics and Google Ads, helps users to design useful data experiments. Additionally, it is possible to identify website areas that need improvement and apply these insights to build customer-friendly and audience-specific website pages.

In this article I will discuss how connecting Google Optimize to Google Analytics can help your marketing team better understand their users’ behavior, and generate analysis that can help improve website engagement and eventually conversions.

Integrate Google Optimize with Google Analytics

Linking Google Optimize and Google Analytics is a two-way data transfer process that brings a series of benefits. According to the official Help Center, when you link Optimize and Analytics:

  • Optimize can calculate experiment results based on the traffic in the view.
  • Optimize can access Analytics goals and data.
  • Optimize 360 customers can target experiments to Analytics audiences.

You can use goals and metrics from Google Analytics as experiment objectives as well as leverage user segments as the target for configuring experiments. Also, you can use experiment dimensions in Google Analytics for every session to create different reports and generate relevant user behavior insights.

Google Optimize offers many reports that showcase the performance of a particular variation against the original. One can assess how variations affect different metrics and goals (objectives). Every experiment can enjoy this real-time reporting, based on the performance of a variation.

However, with the two-way linking process, it is possible to view experiment reports in Google Analytics as well. Select ‘View report in Google Analytics’ in the Google Optimize information panel or browse to Reports > Behavior > Experiments in Google Analytics for accessing reports.

Google Optimize populates three dimensions that can be used as a secondary dimension, in custom reports or to create user segments:

  1. Experiment ID - a unique experiment id that can be seen in the information panel in Google Optimize.
  2. Experiment Name - the experiment name provided by users when creating one in Optimize.
  3. Experiment ID with Variant - the variant number inside a particular experiment in the format {Experiment ID: Variant Number}.

Leveraging these dimensions with other Google Analytics dimensions will get a deeper insight into user behavior. The dimensions will help establish patterns of customer interaction and web experience.

Let us look at two use cases to get a better understanding of customer behavior and learn how to use experiment dimensions more effectively.

1. Applying experiment segments to Shopping Behavior Report

Consider a scenario where an eCommerce company wants to study users’ behavioral patterns through their sales cycle. They should start by implementing and reviewing the comprehensive Shopping Behavior Analysis in Google Analytics ecommerce reports.

The report in the screenshot below shows that many users drop-off from the "Add to Cart" stage, and this drop-off is a significant reason leading to fewer conversions.

Shopping Cart Dropoff Funnel

After a careful analysis, it was found that the most important cause for this drop-off is the addition of shipping costs on the Cart page. Hence, the company decided to run an A/B test to display shipping costs at the Product Details stage itself.

After running the experiment for a certain period, it showed that the original version of Add to Cart outperformed the variant, with the original one still leading to higher conversions.

To understand the reason, they reviewed the Google Analytics Shopping Behavior Report and segmented a set of users who were a part of this experiment and fall under the variant version. They applied a condition on Experiment ID to uniquely identify such users, as shown in the screenshot below:

Experiment ID Analytics Segment

Applying this segment suggests that the drop-off from the Add to Cart step decreased to a great extent. But the Checkout phase still forces many users to exit the shopping funnel (see report below). This could be due to the time-consuming process of filling the checkout form.

Optimize Funnel Analytics

Based on this, they decided to run another A/B test, redesigning the checkout form with precise instructions and uncluttered space. As a result, the new variation improved the checkout process and even boosted conversions significantly.

2. Applying experiment segments to All Channels Report

It is very important for a lead generation business to measure the performance of their lead form. A seamless form experience is critical to their conversion rate.

In the scenario below, a business wanted to improve the form drop-off rate and boost conversions with the help of an A/B test. They decided to redesign their form to remove the contact number field that led several users to exit the form without completing it.

Form Redesign AB Test

The experiment results show that both versions performed well, and there was no clear winner. For further analysis, they reviewed the Channels Google Analytics report (to be found under Acquisition > All Traffic > Channels). They segmented out the users who were a part of this experiment with the help of an ‘Experiment ID’ segment condition to include only ‘Experiment IDs’ with a variation as a secondary dimension to assess the difference in user behavior. Below you will find screenshots of both the segment and the resulting report.

Optimize Test Segment

Optimize Test Analytics Results

The original version of the lead form was a winner for users arriving on the website through Direct traffic, but the variant performs better for Paid Search. Users coming from Paid Ads might not be aware of the brand or might need some time to build trust in the brand to share personal details. But users landing directly on the website know the brand, and hence they are keen on sharing personal details more easily.

Conclusion

Google Optimize is great for small-to-medium size businesses who want to optimize conversions. The tool relies on the experiment results to optimize campaigns and other marketing strategies. Google Optimize and Google Analytics include a two-way integration process, which allows users to run experiments on targeted users (audiences from GA) and goals already defined in Google Analytics, along with the study of these user segments for their website visits back in GA.

Citation: Feature Image Attribution

Sharing is caring!