Custom Audiences

What This Page Answers

Custom audiences are advertiser-owned or advertiser-defined audiences created from sources such as website visitors, customer lists, app users, video viewers, lead forms, purchases, or CRM data. They are useful for retargeting, exclusions, lookalikes, lifecycle strategy, and conversion-quality feedback. But a custom audience is only as useful as the data source behind it.

What Custom Audiences Are For

Custom audiences can support:

Retargeting warm users.
Excluding existing customers from prospecting.
Building lookalike or similar audiences.
Segmenting leads by CRM stage.
Re-engaging app users.
Sequencing creative by funnel stage.
Separating new vs returning customer reporting.
Sending qualified lead or purchase feedback into platform learning.

Common Audience Sources

Source	Best Use	Main Risk
Website visitors	General retargeting	Too broad; low intent mixed with high intent
Product viewers	Product-specific retargeting	Can over-credit returning shoppers
Cart abandoners	High-intent recovery	Small audience; frequency risk
Purchasers	Exclusions, upsell, lookalikes	Needs clean deduplication and recency logic
CRM leads	Nurture, exclusions, quality feedback	Data freshness and match rate issues
High-LTV customers	Lookalike source	May be too small if filtered too tightly
Video viewers	Mid-funnel retargeting	Attention does not always equal buying intent

Quality Requirements

Custom audiences are only as good as their source data. Check:

Pixel or server event quality.
Consent and privacy permissions.
CRM freshness.
Audience size.
Event deduplication.
Exclusion logic.
Recency windows.
New vs returning customer separation.
Whether the audience represents value or just activity.

How To Choose Audience Windows

Short windows are higher intent but smaller. Long windows are larger but less intent-specific. Example approach:

1-7 days: high-intent retargeting or cart recovery.
8-30 days: consideration and objection handling.
31-180 days: broader nurture, exclusions, or reactivation.
Purchasers: separate by recency, value, and lifecycle stage.

The correct window depends on purchase cycle length. A $20 impulse product and a $20,000 B2B purchase should not use the same audience logic.

Common Mistakes

Using all website visitors as if they are high intent.
Not excluding purchasers from acquisition campaigns.
Building lookalikes from low-quality leads.
Letting old CRM lists decay.
Treating video views as purchase intent.
Using custom audiences without checking match rate or size.
Over-retargeting small audiences and creating attribution inflation.

Practical Rule

Custom audiences should improve decision quality. If they only make reporting look better, they may be creating attribution inflation.