How to Do Predictive Analysis in Google Analytics

Knowing what your website visitors did yesterday is useful, but what if you could predict what they'll do next week? That’s the promise of predictive analysis in Google Analytics 4. This guide will show you how to find and use GA4’s built-in predictive features to identify which users are likely to buy, which are about to leave, and what kind of revenue you can expect, so you can tailor your marketing for what’s coming next.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

What Exactly is Predictive Analytics?

Predictive analytics uses historical data, statistical algorithms, and machine learning techniques to identify the likelihood of future outcomes. In simple terms, it's about making educated guesses about the future based on the past.

Traditional web analytics is descriptive. It tells you what happened:

• How many people visited your site yesterday.
• Which blog post got the most views.
• What your conversion rate was last month.

Predictive analytics, however, is prescriptive. It helps you anticipate what will happen:

• Which first-time visitors are most likely to make a purchase in the next 7 days.
• Which existing customers are at risk of not returning.
• How much revenue your highest-spending users are likely to generate next month.

This shift from looking backward to looking forward is a game-changer for marketers. It allows you to move from reacting to results to proactively shaping them.

Welcome to GA4: Where Predictive Analytics is Built-In

For years, this kind of forecasting required data scientists, expensive software, and complex models. Universal Analytics (the old Google Analytics) was purely a reporting tool, it didn’t have any native predictive capabilities.

Google Analytics 4 is fundamentally different. It uses Google's machine learning models to analyze your data and automatically generate predictive insights. The goal is to make this powerful technology accessible to everyone, not just data experts. GA4 looks for patterns in user behavior - things like specific events they trigger, pages they visit, or how frequently they engage - to calculate the probability of them taking a certain action in the future.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

What You Need to Get Started with Predictive Metrics

Before you can use predictive analytics in GA4, your account needs to meet specific data eligibility requirements. Google has these thresholds in place to ensure its machine learning models have enough data to make reliable predictions. If you don't have enough data, the predictions won't be accurate.

Here are the key requirements:

1. Correct Event Tracking

Your predictive metrics are anchored to specific events. To get started, you must be tracking the purchase event (and/or the in_app_purchase event for apps). For e-commerce sites, this means having your GA4 ecommerce tracking set up properly. The model needs to see who is buying and who isn't to learn what separates them.

2. Sufficient Data Volume

This is usually the biggest hurdle for smaller sites. To unlock the metrics, you need to meet minimum user and event counts over a recent 28-day period:

• For Purchase Probability and Predicted Revenue: You need at least 1,000 users who triggered a purchase event and at least 1,000 users who did not. Think of it as having 1,000 "successful examples" and 1,000 "unsuccessful examples" for the model to learn from.
• For Churn Probability: You also need at least 1,000 users who triggered a purchase or in_app_purchase event and at least 1,000 users who did not.

3. Sustained Performance

Meeting these thresholds isn't a one-time thing. Your account must maintain this level of traffic and conversions consistently. If your user or event count dips below the minimum for an extended period, the predictive models will become unavailable until the thresholds are met again.

Once you meet the criteria, it can still take a little time for the metrics to become available in your account. You can check your eligibility by navigating to the "Audiences" section in GA4 and looking for the "Predictive" audience templates. If they're available, you're good to go!

GA4's 3 Core Predictive Metrics

GA4 provides three primary predictive metrics that you can use to build audiences and gain insights.

1. Purchase Probability

What it is: This metric predicts the likelihood that a user who has been active in the last 28 days will trigger a purchase event within the next 7 days.

How to use it: Purchase Probability is perfect for identifying high-intent users who are on the verge of buying. You can create an audience of users with the highest probability and target them with remarketing campaigns on Google Ads, offering a small discount or free shipping to nudge them across the finish line.

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

2. Churn Probability

What it is: The flip side of purchase probability. This predicts the likelihood that a recently active user (in the last 28 days) will not be active on your site or app within the next 7 days.

How to use it: Churn Probability helps you spot at-risk customers before you lose them for good. You can build an audience of users with a high probability to churn and run targeted retention campaigns. This could be an exclusive offer, a "we miss you" email, or a survey to gather feedback.

3. Predicted Revenue

What it is: This metric forecasts the total revenue you can expect from all purchase conversions within the next 28 days from a user who was active in the last 28 days.

How to use it: Predicted Revenue is excellent for identifying your potential highest-value customers. You can create an audience of these "future VIPs" and give them extra attention, enroll them in loyalty programs, or use them as a seed list to create powerful lookalike audiences in Google Ads to find more customers just like them.

How to Create and Use Predictive Audiences in GA4

Once your property is eligible, creating these audiences is straightforward. They are designed to be easily accessible from the Audience builder.

Step-by-Step Guide:

1. Navigate to the Admin section of your GA4 property (the gear icon in the bottom-left).
2. In the 'Property' column, click on Audiences.
3. Click the blue New audience button.
4. On the next screen, you’ll see a section for "Suggested audiences." Click on the Predictive tab.

If you're eligible, you'll see a list of pre-configured audience templates, such as:

• Likely 7-day purchasers
• Likely 7-day churning users
• Predicted top spenders in 28 days
• Likely first-time 7-day purchasers

Let’s say you want to build an audience of your most likely buyers. You would click on "Likely 7-day purchasers."

From here, you can configure your audience. The main setting is a slider that lets you define the user percentile range you want to include. For example, you can choose to include an audience of users who fall between the 80th and 100th percentile for purchase probability. This targets the top 20% of users most likely to convert.

You then set the membership duration (how long a user stays in the audience) and give your audience a name, like "High-Intent Prospects - Top 20%".

Graphed

Still Building Reports Manually?

Watch how growth teams are getting answers in seconds — not days.

Watch the Demo

Practical Audience Examples to Build:

• The Hesitant Shopper: Users with a moderate-to-high purchase probability who have added items to their cart but not completed a purchase.
• The Quietly Leaving Customer: Users in the 75th-95th percentile for churn probability. They aren’t your most at-risk yet, but they're showing signs of slipping away.
• The Potential Whale: An audience of the top 5% of users with the highest predicted revenue.

Connecting Your Predictive Audiences to Google Ads

Creating these smart audiences is only the first half of the equation. Their real power is unleashed when you connect them to Google Ads for targeted campaigns. To do this, make sure your GA4 property is properly linked to your Google Ads account.

Once linked, your predictive audiences will be automatically shared with Google Ads. You can then use them in your campaigns for focused targeting and creative messaging:

• Remarket to "Likely Purchasers": Show these users conversion-focused ads with compelling offers to encourage an immediate purchase.
• Re-engage "Likely Churners": Run ads with a reminder of your brand's value or introduce a new product to pique their interest again.
• Build Lookalike Audiences: Use your "Predicted Top Spenders" audience as a base for Google's lookalike (similar) audiences to find new users who share traits with your best potential customers.

By focusing your ad spend on segments defined by future behavior rather than just past actions, you can significantly improve your campaign efficiency and return on investment.

Final Thoughts

Google Analytics 4's predictive capabilities provide marketers with powerful tools to anticipate customer behavior. By moving from a reactive to a proactive mindset, you can build smarter audiences, personalize your marketing efforts, and ultimately drive better business outcomes. Getting your events set up correctly and meeting the data thresholds is the first step, but the insight you gain is well worth the effort.

Analyzing predictive metrics in GA4 is incredibly valuable, but it's often just one piece of your full marketing picture. To get real answers about what's driving results, you usually need to combine this data with information from your ad platforms, your Shopify store, and your CRM. We built Graphed to solve this challenge. Instead of toggling between a dozen tabs, you can connect your data sources in one place and ask questions in plain English to instantly get dashboards and reports. This turns hours of manual report building into simple, 30-second conversations about your business performance.