Set up audiences

Learn how to define groups of visitors you can use in your optimizations.

Set up audiences

Learn how to define groups of visitors you can use in your optimizations.

Audiences in Optimize

An audience is a group of visitors that matches criteria you define.

You can use audiences to make sure an optimization reaches the right visitors. For example:

  • Mobile visitors
  • Returning visitors
  • Visitors from a specific campaign
  • Visitors from enterprise companies
  • Existing customers
Create audience modal with a name field and rules for device, visitor, location, time, campaign, company, and behavior.

Once you create an audience, you can reuse it across optimizations.

How audiences work in optimizations

Audiences play different roles depending on the kind of optimization you’re running.

Traditional Test and AI Optimize use audiences the same way: to define who can enter the optimization. Manual personalization uses audiences to decide which experience each visitor sees.

Audience = who can participate

For traditional tests, you'll use an audience to limit the test to a specific group of visitors.

Example: Test a higher homepage CTA only with mobile visitors.

Eligible visitors enter the test, then traffic is split across the variations.

Audience targeting splits 4,200 mobile visitors evenly between base and Variation A; 5,800 other visitors see the original page.

Audience = who can participate

AI Optimize uses audiences the same way as a traditional test: the audience defines who is eligible for the optimization.

Example: Run several headline variations only for paid campaign visitors, then let AI decide which variation each eligible visitor sees.

The audience controls eligibility. AI Optimize controls variation delivery.

AI Optimize adapts delivery of base and three headline variations for 2,800 paid campaign visitors; 7,200 others see the original page.

Audience = who sees what

In a manual personalization, audiences are assigned to individual variations.

Example: Show enterprise visitors an enterprise-focused messaging and small-business visitors SMB messaging.

Here, the audience directly determines which experience a visitor receives.

Audience rules show tailored variations to 1,500 enterprise and 3,200 small-business visitors; 5,300 others see the original page.
Best practice

Plan for audience overlap

In a manual personalization, a visitor may qualify for more than one audience. For example, someone might be both an Existing customer and part of an Enterprise company.

When that happens, audience priority determines which experience they receive. Put the more specific or important audience higher in the priority order.

Manage audiences & variations modal with draggable rows for setting audience priority: Returning visitors, New visitors, then Agriculture: Crops.

Choose how to define your audience

Most audiences in Optimize are rules-based. You define one or more conditions, and visitors who match those conditions become part of the audience.

The rules you can use come from several kinds of visitor data.

Start with built-in visitor signals

Optimize includes built-in signals you can use to define an audience without connecting another data source. These can include things like a visitor’s device, location, referral source, or previous activity.

For example, you could create an audience of returning mobile visitors or visitors from a specific region.

Build more specific audiences with connected data

On Enterprise plans, you can use additional data sources to create more specific audience rules:

Campaign & URL data

Campaign and URL data are details passed through a visitor’s URL, such as UTM parameters.

Example:

‍utm_campaign = enterprise-launch

CRM & marketing data

Connected CRM and marketing tools can make lead, customer, or account information available for audience rules.

Example:

Existing customers

Company data from Enhanced Match

Enhanced Match can provide company-level information for some otherwise anonymous visitors, such as industry, company size, domain, or estimated revenue.

Example:

Large healthcare companies

Custom attributes

Custom attributes let your team pass additional visitor information into Optimize when the built-in signals don’t cover what you need.

Example:

‍customer_tier = enterprise

Code-based audiences

For advanced cases that rules can’t express, you can build a code-based audience using custom JavaScript.

The code evaluates whether a visitor meets the condition. If it returns true, the visitor matches the audience.

For example, you could create an audience for visitors whose browser window is at least 760px tall, so a large image is only shown when there’s enough room to see it on page load.

Create audience modal after selecting “Code Your Audience,” with JavaScript defining “Tall screens” as a viewport at least 760 pixels tall.
760px or taller → true → visitor matches the audience

For most targeting needs, start with rules-based audiences. Use code-based audiences when the targeting logic you need isn’t available through rules.

Start simple

Audience rules can get specific quickly. But the more conditions you add, the fewer visitors will qualify.

A four-step audience graphic showing all visitors narrowing to mobile visitors, then returning mobile visitors, then returning mobile visitors from Campaign A, with fewer visitors at each step.

Start with the simplest audience that supports your goal. Add more rules only when they serve a real targeting need.

Before using an audience, check that it:

  • Includes the visitors you intend
  • Excludes the visitors you don’t
  • Still has enough traffic to support the optimization

For tests and AI Optimize, that last point matters especially because smaller audiences mean less data to learn from.

Ready to keep going?

Next, you’ll use the foundation you’ve set up — tracking, connected data, goals, and audiences — to find high-impact opportunities worth optimizing.

1

Intro

Coming soon

1

Welcome & overview
2:00
Welcome & overview
Coming soon

1

Intro to Optimize
2:00
Intro to Optimize
Coming soon

1

Optimization types
6:22
Optimization types
Coming soon

2

Setup & configuration

Coming soon

2

Tracking & consent
2:30
Tracking & consent
Coming soon

2

Connect your data
3:00
Connect your data
Coming soon

2

Set up goals
4:00
Set up goals
Coming soon

2

Set up audiences
4:30
Set up audiences
Coming soon

3

Deciding what to optimize

Coming soon

3

Finding high-impact opportunities
3:00
Finding high-impact opportunities
Coming soon

3

Build an optimization roadmap
2:00
Build an optimization roadmap
Coming soon

3

Common things to test and personalize
3:00
Common things to test and personalize
Coming soon

4

Building and launching optimizations

Coming soon

4

Traditional tests
3:30
Traditional tests
Coming soon

4

Manual personalizations
3:30
Manual personalizations
Coming soon

4

AI Optimize for tests and personalization
3:30
AI Optimize for tests and personalization
Coming soon

5

Reviewing results and monitoring performance

Coming soon

5

Traditional test reports
5:00
Traditional test reports
Coming soon

5

Manual personalization reports
4:30
Manual personalization reports
Coming soon

5

AI Optimize reports
4:30
AI Optimize reports
Coming soon

5

Monitoring site performance
3:30
Monitoring site performance
Coming soon

6

Wrap up

Coming soon

6

Optimization best practices
4:30
Optimization best practices
Coming soon

6

Next steps & additional resources
2:00
Next steps & additional resources
Coming soon

Course progress

0%

Assessment

Up next

Finding high-impact opportunities

Use your data to find high-impact opportunities and form a testable hypothesis.
Continue to next lesson
→
Complete course
✓