How Google Analytics Serves as an Advertising Analytics Tool

How Google Analytics Serves as an Advertising Analytics Tool

Trapica Content Team

Industry News
6 min read
April 29, 2021

For the longest time, we’ve recommended Google Analytics for all businesses aiming to learn more about their audience. With knowledge and understanding, we’re in a better position to create ads and content that resonate and generate results. Thanks to the machine learning systems Google uses, we can make predictions for the future and shape our strategy around these predictions. Now, Google is moving forward again with predictive audiences and predictive metrics features for App + Web properties. With that said, it sadly isn’t available for regular Analytics uses. The two features include:

  • Churn Probability - How likely is it that somebody who was recently active WILL NOT visit your app or website over the next week?
  • Purchase Probability - After visiting your website or app, how likely is it that the consumer WILL spend their money and make a purchase?

As you can imagine, these two changes are huge for both businesses and marketers. On the one hand, we can retain those who are less likely to return or become a customer. On the other, there’s an opportunity to optimize our marketing efforts for those more likely to convert.

How Do the Features Work?

As you scroll through Google Analytics, you’ll see ‘Likely 7-day purchasers’. This uses Purchase Probability to discover the group of people most likely to purchase within the next week. The page will also show ‘Likely 7-day churning users’ which has the group of people unlikely to make a purchase in the next seven days.

Building Audiences

Is this an important feature? Yes, because previously, reaching out to an audience likely to purchase would take lots of guesswork. For example, you might have created a campaign to target those who had abandoned a shopping cart. What about the people who didn’t add products to their cart but still have an interest in buying? Some people add products to a basket with no intention to buy, and others don’t add products to a basket despite an intention to buy. With these new features from Google, we hope this will bridge the gap and allow for more accurate marketing efforts.

If you have a website and an app, the benefit of this type of predictive audience is that you can target both. Using Google Analytics, you’ll see a list of names most likely to convert over the next week. Rather than wasting time trying to cover a larger list, you can hone in on the most valuable prospective customers and hit them with a personalized Google Ads campaign.

Just because there are people unlikely to convert in the next week doesn’t mean you should disregard them and generate a new audience. Instead, run a second Google Ads campaign and guide them towards a detailed article. Perhaps you can explain their problem and your solution, transforming them into people who may purchase in the next seven days.

Analyzing Audiences

It’s not all about building audiences, and Google believes the two metrics will come in handy for those looking to analyze data. As an example, you might currently use the User Lifetime technique which shows the number of users generated from each marketing campaign. With this new feature, you can see the marketing campaigns that convert the most people from the Purchase Probability list. If one campaign seems to convert lots of people on this list, you can pump more of the budget into this campaign, while also using it as a learning point for other campaigns.

Using the Feature

At the moment, the new feature is only available to some users and is currently in a beta testing phase. To gain access, you will need to create an App + Web property. As long as specific thresholds are met, the feature will appear for those who are measuring in-app purchases automatically, and those who have implemented purchase events.

In a recent survey from Statistica, only 20% of major businesses fail to utilize the power of predictive analytics. For the vast majority, they see the strength of this automation and spend a section of their marketing budget on the technology every year.

This update will help everybody, but especially small businesses with little experience in this area. For marketers who work on behalf of these small businesses, the feature provides assistance and should make your job easier. Thanks to the update, you’ll get suggestions for a predictive audience that you can then take forward into the Audience Builder.

If we refer back to the question of this article, can the new Google Analytics feature predict your audience? No tool will ever have the ability to predict the future and precisely guess the customers that will convert and the customers that won’t. Even with these new features, there will be people from the ‘unlikely to convert’ group who purchase within seven days. Likewise, you will find people likely to convert who don’t purchase. However, Google is now pushing the boundaries of predictive analysis; it can predict our audience with more accuracy than ever before.

We now have this combination of predictive audiences and predictive metrics, and it’s a huge step forward for the utilization of a marketing budget. If the feature is successful, a business can convert more leads with every dollar spent.

Creating a Predictive Audience

If you’ve been into the Audience Builder, you’ll know that it all starts with a selection of suggested predictive audiences. For construction material services with an app and website, for example, they can use this new feature as part of an effort to increase the number of online sales made each month. When assessing analytics, it will now suggest an audience of people likely to convert in the next week.

Assuming your property is eligible for the feature, suggested audience templates are useful to make an audience based on the predictions. With this, we can create a tailored message for the people on the cusp of purchasing. Follow the steps below to get started:

  • Choose your App + Web property once logged into Analytics
  • Select Audiences and then ‘New audience’
  • Choose Predictive under the ‘Suggested audiences’ option  
  • Some of the templates will be ready to use and this is true when the prerequisites are met for the suggested predictive audiences
  • Adjust the template as required and add any non-predictive conditions you need

Using Predictive Metrics

To predict the actions of future customers, one of the most valuable resources we have is the history of existing ones. One of the best techniques for analyzing customer activity is the user lifetime technique, touched on previously. Essentially, the goal is to assess user behavior throughout their interaction with the brand. There are three main insights with the user lifetime technique.

Firstly, you learn the campaigns that are generating leads with the highest potential value. With the new prediction models, this means customers with low churn probability and high purchase probability.

Secondly, you learn the campaign or source that generates leads with the most lifetime revenue. Often, this is compared with revenue over a given month.

Thirdly, you pick up all sorts of individual insights. For instance, one of our favorites is the most recent purchase from monthly active users. Alternatively, those driving interest in an app might be more interested in when a monthly active user last used the app.

Importance of the New Feature

Should you pay attention to this new Google Analytics feature? We believe so, and there are many reasons why. Primarily, if you have App + Web properties, the beta will help to both analyze marketing spend and build audiences. With Google Ads, it’s easy to wonder just how your marketing budget is being spent. We welcome features that improve the efficiency of marketing spend, and this is just another one from Google.

Over the years, some businesses have found it almost impossible to determine the readiness of customers to spend their hard-earned money. While cart abandonment is certainly a good indicator, this can still lead businesses awry. If we look at Facebook, there’s not much to go on in this area. We could consider lookalike audiences and interests, but this sometimes isn’t enough to determine whether or not a prospective customer will spend money within the next week. We can optimize for conversion, but we expect this new Google feature to finally provide competition when trying to achieve this goal.

For media buyers, they don’t need to rely on optimizing for conversions on Facebook. If possible, we recommend signing up for the App + Web property beta as soon as possible. Play around with the metrics and predictive audiences to learn how they could benefit your business. Also, take advantage of Google resources because the company explains how to make the most of the new feature. Remember, build audiences, and analyze existing customers to predict your audience.

If all goes well, you can personalize a message for those most likely to convert while reaching out to those unlikely to convert with valuable content - you win either way!

Learn how Trapica can help you automate and optimize your Google ad campaigns

Industry News
6 min read
April 29, 2021