Audience Targeting

Paid Media

Also: Ad Targeting · Audience Segmentation · Target Audience

TypesInterest, behavioural, custom, lookalike, retargeting

Watch outOverly narrow targeting inflates CPMs

Start withCustom audiences from your own data

Better signalPurchaser lookalikes beat interest audiences

Quick definition

Audience targeting is the process of defining who sees your paid advertising based on characteristics like demographics, interests, behaviour, past interactions or similarity to existing customers. Better targeting means your budget reaches people more likely to respond, improving conversion rates and reducing cost-per-acquisition.

How it varies across Australia

Custom audiences built from first-party data (email lists, website visitors, purchasers) consistently outperform interest-based targeting for Australian advertisers. Lookalike audiences built from purchaser lists typically achieve 20-40% lower CPA than broad interest targeting. Australian advertisers with CRM data they're not using for targeting are leaving significant efficiency gains on the table.

See acquisition performance benchmarks →

Custom audiences

Audiences built from your own data: email lists, website visitors, app users, purchasers. These are your highest-quality targeting inputs because they're based on real interactions with your business.

Lookalike audiences

Platform-generated audiences that resemble your custom audience. A 1% lookalike of your purchasers finds the people most similar to people who have already bought from you. Quality of the seed audience determines quality of the lookalike.

Interest targeting

Reaching people based on their stated or inferred interests. Broad, lower-fidelity than custom or lookalike audiences. Useful for awareness campaigns and when you have limited first-party data.

Retargeting

Showing ads to people who have previously interacted with your brand: website visitors, video viewers, social engagers. Higher purchase intent than cold audiences.

What it actually means

Audience targeting decides who sees your paid advertising. Every campaign you run has an implicit or explicit audience definition, and the quality of that definition directly affects how efficiently your budget converts to results.

The hierarchy of targeting quality, from best to worst, generally runs: retargeting your purchasers, retargeting your warm audiences (website visitors, email subscribers), lookalike audiences built from purchasers, lookalike audiences from website visitors, interest-based targeting, and broad demographic targeting.

This hierarchy exists because the further you move from actual customer data, the more you're relying on the platform to infer intent. Interest targeting on Meta is based on pages people follow, engagement behaviour and demographic patterns. It's an educated guess. A purchaser list is proof.

Australian businesses often underinvest in custom audience infrastructure because building it requires clean CRM data and a willingness to connect first-party data to ad platforms. The businesses that do this consistently see materially better paid media efficiency.

The best targeting input you have is a list of people who already bought from you.

How it shows up

Audience quality shows up in CPA, ROAS and CTR by audience segment. Separate your reporting by audience type to compare performance. Custom and lookalike audiences from high-quality seeds should show lower CPAs than interest audiences. If they don't, check the seed audience quality or consider whether your retargeting pools are too small.

The Australian context

Australia's smaller market creates audience size constraints. A 1% lookalike of 500 purchasers may only generate an audience of 20,000-30,000 on Meta in Australia. This can limit scale while maintaining efficiency. The practical response is broader lookalike percentages (2-5%) for scale with tighter creative targeting for efficiency.

Where people get this wrong

Audience targeting that's too narrow.Audiences under 30,000 on Meta create competition for impressions against yourself, inflating CPMs significantly. In Australia, narrow targeting gets expensive fast. Broaden gradually and use creative differentiation rather than audience narrowing.

Not segmenting retargeting by recency and intent.A 7-day website visitor and a 90-day website visitor have very different purchase intent. Separate retargeting audiences by recency and use different creative and offers for each.

Uploading a full email list as a lookalike seed.If your email list contains subscribers from five years ago who never purchased, you're including low-quality signals in your seed. Lookalike quality is a direct function of seed quality. Use purchasers or high-engagement subscribers as your seed.

Common questions

What is the best audience targeting strategy for Facebook ads?

Start with retargeting campaigns to warm audiences (website visitors, email subscribers, past purchasers). Then build lookalike audiences from your purchaser list. Use interest targeting as a third layer for broader awareness. Prioritise your own data over platform-inferred interests.

How large should a lookalike audience be?

In Australia, 1% lookalikes are typically 100,000-300,000 people depending on the seed. For small seed audiences (under 500 people), a 2-3% lookalike may perform better because the algorithm has more flexibility to find relevant patterns. Test different percentages.

How do I improve audience targeting performance?

Clean and grow your first-party data. Upload purchaser lists, segment by purchase recency and value, and build lookalike audiences from your best customers rather than your full contact list. Better seed data produces better lookalikes.

Keep exploring

About New Rebellion

New Rebellion is a marketing intelligence consultancy. We build tools, score Australian businesses on how their marketing actually performs, and publish Debrief every day. This dictionary is part of how we work in the open.

How we think →