Meta AdsModule 2: Audience and deliveryLesson 4 of 11
Course progress27%

18 min lesson · Updated August 2026

How do Meta audiences work?

Meta audiences combine advertiser controls, suggestions, first-party/custom data and delivery predictions; many inputs guide the system rather than defining a perfectly fixed group.

What you will learn

By the end, you will understand:

  • Distinguish broad, detailed, custom and lookalike audiences
  • Separate strict controls from Advantage+ suggestions
  • Use audience data with lawful permission and realistic expectations

Visual explainer

See the idea clearly.

An audience is an eligible opportunity pool

Audience settings help Meta decide who may receive or be prioritized for an ad. Delivery still depends on the objective, auction, creative, placements, budget and predicted action. An audience estimate is not a guaranteed number of unique people who will see or buy.

Broad targeting gives the system room to learn within essential controls. Detailed targeting uses interests, demographics or behaviors available in the account. Availability and granularity change over time and by country or policy.

Advantage+ audience uses controls and suggestions differently

With Advantage+ audience, inputs such as interests, age/gender suggestions, custom audiences and lookalikes may guide initial delivery while the system can search more widely when it predicts better results. Meta describes location, minimum age, language and eligible custom-audience exclusions as controls that can remain strict.

Read the exact interface language. A suggestion is not a boundary. If the business has a legal, safety or service-area limit, use the supported control and verify actual customers.

Control

A supported boundary the system should not expand beyond, such as selected location or minimum age.

Suggestion

A signal describing likely customers that Advantage+ delivery may move beyond.

Custom Audiences reconnect known relationships

Custom Audiences can be created from eligible customer lists, website/app activity, offline events, Meta engagement and other supported sources. They can support retention, exclusion, sequential messaging or seed audiences.

The source must be lawful and policy-compliant. A purchased list or scraped contact data is not made acceptable by hashing. Respect user choices and delete or update data when required.

Lookalikes model similarity, not identity

A Lookalike Audience helps find new people who share patterns with a selected source audience. Source quality matters: verified high-value customers are usually more informative than all email subscribers mixed together.

Similarity does not mean the new person has the same needs or will convert. Country, source size, recency and objective can affect usefulness. In Advantage+ setups, a lookalike may act as a suggestion rather than a strict target.

Creative also selects the audience

A video about managing payroll speaks to different people from a broad “grow your business” slogan. Clear creative can let relevant viewers recognize themselves and unsuitable viewers move on.

Do not use personal-attribute language that implies knowledge of sensitive traits. Meta policies restrict statements or targeting around personal attributes and sensitive categories.

Overlap and exclusions need a purpose

Multiple ad sets can reach some of the same people. Overlap is not automatically harmful, but fragmented campaigns can compete for limited evidence and make interpretation difficult. Consolidation can help when the objective and creative learning are shared.

Use exclusions for real strategy—for example excluding recent purchasers from an acquisition offer—not as a ritual. Some Advantage+ campaign types handle existing-customer controls through specific settings.

Audience review

  • Which settings are true controls and which are suggestions?
  • Does the location match the service or fulfilment area?
  • Does the minimum age meet product and legal requirements?
  • Is the source audience accurate, current and lawfully collected?
  • Are sensitive or prohibited attributes avoided?
  • Does creative clearly qualify the intended person?
  • Are recent customers included or excluded intentionally?
  • Do CRM/order records confirm who actually responded?

Real-world example

Example: premium children’s furniture

Example

The campaign uses serviceable countries and minimum purchaser age as controls. Verified past purchasers form a lawful Custom Audience and high-value seed. Their lookalike and design interests guide delivery, but Advantage+ can expand beyond suggestions. Creative shows room size, materials, price range and delivery, helping suitable parents self-select.

Try this

Label controls and suggestions

Open one ad set and list every audience input in two columns: strict control or delivery suggestion. For each, write the business reason and the evidence used to verify it.

Common questions

Questions beginners ask.

What is broad targeting?

Using essential controls while giving Meta wider room to find likely responders from objective and creative signals.

What is detailed targeting?

Available interests, demographics or behaviors used to describe likely audiences, subject to platform and policy limits.

Can Advantage+ audience go beyond interests?

Yes. Meta describes many inputs as suggestions that can expand, while supported controls remain boundaries.

What is a Custom Audience?

An eligible audience built from a business relationship or interaction source such as a lawful customer list, website event or Meta engagement.

What is a Lookalike Audience?

A modeled group of new people with similarities to a selected source audience.

Can I upload purchased email lists?

No responsible strategy should do so. Meta requires necessary rights, permissions and lawful basis for customer-list data.

Does narrow targeting guarantee relevance?

No. Available attributes can be inaccurate or incomplete, and creative/offer still determine response.

Should existing customers always be excluded?

No. It depends on acquisition, retention, cross-sell and message strategy. Make the choice intentionally.

Assessment

Check what you understood.

5 questions · instant explanations

1. What is often a strict Advantage+ audience control?
2. What makes a strong lookalike source?
3. What does hashing a customer list prove?
4. How can creative help audience quality?
5. True or false: every audience suggestion is a hard delivery boundary.

Sources

Primary references.