Audiences are groups of your users who share the same attributes. You can share audiences with the advertising products you use, like Google Ads, so you can market to specific groups of users.
What is an audience?
In Google Analytics, an audience is a group of users from your site and/or app who have generated similar behavioral data or who share demographic or other descriptive data (e.g., same age group, same gender, were acquired by the same campaign).
For example, you might want to create an audience of all users who have made a purchase of any kind (purchase event_count > 0).
You might find this audience too large for the practical purposes of your ad campaigns. In this case, you can identify smaller groups of users who have more specific behavior and demographics in common, like:
Users from California who have purchased 1-5 items
Users from San Francisco, California who have purchased 1-5 items in the last 7 days
Users from San Francisco who purchased 1-5 items in the last 7 days and who spent more than $100
You can build audiences around any of the data you collect and you can define your audiences as broadly or narrowly as necessary: All users who purchased sunglasses or All users who purchased the black wraparounds with the blue reflective lenses. Anytime a user visits your site or app and their behavior and/or descriptive data match your audience conditions, then that user is added to the audience. The rate at which an audience accumulates users depends on how often users who visit your site or app engage in behavior and/or match descriptive data that satisfy your audience definition. Audiences with broader definitions (fewer conditions) will accumulate users more quickly than audiences with more narrow definitions (more conditions).
Analytics backfills audiences with up to 30 days of data, if that data is available. If you've already had users who fit your audience definition during the last 30 days, then your audience will be populated with users right away.
What do you do with an audience?
You can use audiences in your reports, in advertising campaigns, and in Optimize experiments.
Reporting
If you do nothing but create the audience, you can open a report about it in Analytics and see how many users from that audience visit your site/app, how many conversions and engaged sessions they generate, and the technologies they use to engage with your content.
In your reports, you can use audiences as the basis for comparisons. (Select Audience name as the dimension and select a specific audience for the dimension value.) Then you can compare audiences in the context of the same data: did my audience of customers from Los Angeles buy more sunglasses than my audience of customers from San Francisco? Who bought more of the really expensive ones?
In the Explore module, you can use the Audience name and Last audience name dimensions to dig deeper into which methods of acquiring specific audiences were more effective, which audiences continued to engage once you did acquire them, how those users navigated through your content and through specific funnels you have in mind.
Advertising
If you have your Analytics property linked to Google's advertising products like Google Ads, Display & Video 360, or Search Ads 360, you share your Analytics audiences with those products.
The Purchasers audience, which Analytics creates by default when your users trigger in_app_purchase or purchase events, is shared with any linked advertising products. When you create the product links between Analytics and your advertising products and select the option to enable personalized advertising, your audiences are available in those other products to use to show users ads based on their specific online behavior (e.g., remarketing ads about products they have viewed on your site but haven't yet purchased). When your audiences are available in Google's advertising products, you can use them to specify who you want to show your ads to. For example, if your audience is All users who purchased sunglasses, then whenever one of the users in that audience opens a site or app on which you're advertising, your ad is automatically entered into the bidding to appear to that user. This lets you spend your ad dollars on users who are most likely to respond to your ads.
Optimize experiments (A/B, multivariate tests)
If you link your Analytics property to Google Optimize, you can use your audiences to target Optimize experiments, for example, showing one version of your site to one half of an audience and another version to the other half to see which version of your site performs better.
When you also have Optimize linked to Google Ads, you can use the same audiences that you export from Analytics to Optimize as the basis for Google Ads campaigns.
Differences between Universal Analytics and Google Analytics 4 audiences
Universal Analytics and Google Analytics 4 don't collect exactly the same data. There are dimensions that are the same in both versions (e.g., Country, City, Age, Gender), but the two versions of Analytics collect different data for other dimensions like source and medium. For example, in Universal Analytics if you use the utm_source parameter to collect data, you get values for the Source dimension; if you use the utm_source parameter in Google Analytics 4, you get values for the Source, First user source, and Session source dimensions. In Universal Analytics, you have the user metric Number of sessions per user; in Google Analytics 4, that metric doesn't exist.
Since a lot of the data you use to define audiences in Universal Analytics and Google Analytics 4 is different between the versions, migrating your Universal Analytics audiences to Google Analytics 4 is currently a manual process of examining your Universal Analytics audience definitions and recreating those in Google Analytics 4 using the data that matches most closely.
Besides the detailed data differences, audiences are essentially the same in both versions: you build them the same way using conditions based on data, like "include users who bought sunglasses" or "include users who put sunglasses in their carts but didn't actually buy them".
One big difference is that GA4 includes predictive audiences, while UA doesn't. See the next section to see what predictive audiences are all about.
Predictive audiences
A predictive audience has at least one condition based on a predictive metric. Analytics generates predictive metrics by modeling the data you collect (if you collect the required minimum amounts of the necessary data) and predicting which of your users will behave a certain way. The predictive models evaluate your data to figure out things like which users are likely to make a purchase, which users are likely to churn (stop using your site/app), and how much revenue users are likely to generate.
Predictive audiences are helpful when you're trying to figure out who to spend your marketing budget on -- anybody who's ever made a purchase or just the users who are more likely to make a purchase in the next 7 days?
If Analytics is generating predictive metrics for your property, you can use them when you build your own audiences; for example, "include users from San Francisco AND users who are likely to make a purchase in the next 7 days".
An additional benefit of Analytics generating predictive metrics is that it uses them to generate predictive audiences like Likely 7-day purchasers and Predicted 28-day top spenders that you can use as-is or as the basis for your own custom audiences.
Comments