What audience refinements does Apple Search Ads support?

Four refinement dimensions: gender (male, female, all), age range (custom bands, e.g., 18-24 / 25-34), device class (iPhone, iPad), and customer type (new users, existing users, all users). Refinements are set at the ad group level and applied as AND filters.

Should I use audience refinements?

Selectively. Most acquisition campaigns benefit from leaving refinements off — Apple's algorithm already optimizes toward likely converters. Use refinements when you have strong LTV evidence that a specific segment monetizes differently — for example, a finance app where 25-44 generates 80% of LTV.

Does enabling audience refinements lower CPI?

Not necessarily. Refinements cut reach, so impression volume drops. CPT may rise (smaller eligible set), and CR may rise (more targeted match). Whether net CPI drops depends on the refinement's alignment with conversion drivers. Test before assuming.

What are Audience Refinements in Apple Search Ads?

Audience Refinements are targeting filters that narrow which users in an Apple Search Ads storefront are eligible to see your ad. They sit at the ad group level and apply after keyword and storefront matching.

Four refinement dimensions are available:

Gender — Male / Female / All
Age range — Custom bands (e.g., 18-24, 25-34, 35-44, 45-54, 55+)
Device class — iPhone / iPad / All
Customer type — New users / Existing users / Returning users / All

Refinements apply as AND filters. Setting “Female AND 25-34 AND iPhone” cuts the eligible audience to users matching all three.

Strong LTV signal

If your app’s analytics show a specific demographic segment generates substantially higher LTV, refining ad groups toward that segment can lift ROAS even at lower volume.

Examples where this works:

Finance apps — 25-44 age range often dominates LTV
Productivity tools — Working professionals (25-54) outweigh younger users
Niche dating apps — Gender-specific apps benefit from gender refinement
iPad-first apps — Stylus-based or design apps that target iPad explicitly

Customer type segmentation

This is the most universally useful refinement. The customer type options:

Customer type	When to use
New users	Primary acquisition campaigns — exclude returning users to keep CPI clean
Existing users	Re-engagement or upgrade campaigns targeting current users
Returning users	Win-back campaigns for users who deleted the app
All users	Default; let Apple’s algorithm decide

Setting “New users only” on Brand campaigns prevents your own existing users from clicking your brand ad (which would be a reattribution rather than a new install).

Re-engagement campaigns

Some operators run dedicated Existing users + Returning users campaigns for subscription apps where retention monetizes downstream. These work in addition to the primary New-users acquisition campaigns.

Broad acquisition

Most acquisition campaigns benefit from no refinement. Reasons:

Apple’s own algorithm already optimizes auction matching toward converters
Refinements cut reach without guaranteed CR lift
Over-refining produces low-volume ad groups that take longer to gather statistically significant data

Small accounts

Accounts under $5K/month spend usually have too little volume to benefit from refinement. The data per refined segment is too thin to draw conclusions.

Discovery campaigns

Discovery is exploration — refining cuts the exploration set. Keep refinements off on Discovery and apply them only after promotion to Exact-match production ad groups.

Strategy	Configuration
Default acquisition	No refinements; let Apple optimize
Brand isolation	Customer type = New users (prevents existing-user re-clicks)
iPad-first apps	Device class = iPad on Brand and Category campaigns
Gender-specific apps	Gender filter on all ad groups
Re-engagement	Customer type = Existing or Returning users in dedicated campaign
Age-targeted vertical	Age range refinement on Discovery + Category

Common failure modes

Over-refining. Setting gender AND age AND device AND customer type cuts reach to a small fraction; volume plummets and impression share drops.
Refining based on belief, not data. “Our app is for women” — verify with backend analytics before refining; many gender-skewed apps actually attract significant other-gender users who monetize well.
Forgetting refinements when re-allocating budget. A refined ad group with $5K budget may not be able to spend the budget because the eligible audience is too small.
Not testing on/off. Refinement decisions should be A/B tested when possible — run two parallel ad groups with same keywords, one refined, one not, observe CR and CPI.

How ASAPilot helps

ASAPilot’s account audit reviews refinement usage and surfaces:

Heavily refined ad groups with low impression share (over-refined)
High-spend ad groups with no customer type filter on Brand (potential leakage)
Refined ad groups where unrefined performance was better historically (un-refine candidate)

See audit guide and pricing.

Ad Group — refinements are set per ad group
Apple Search Ads — the platform
Storefront — refinements apply within a storefront