How US and UK benchmarks read honestly

Headline medians, honestly labelled#

Start with the headline medians, each labelled as estimated vendor data. SplitMetrics 2025 prints CPT near 2.25 dollars, CPA near 3.76 dollars, tap through rate near 9.7 per cent and conversion rate near 66.2 per cent, per the SplitMetrics benchmarks. AppTweak prints a global median CPT near 0.92 dollars against a US median near 1.91 dollars, per the AppTweak benchmarks.

Apple own guidance puts top position conversion above 60 per cent, which frames why brand defence converts so fiercely. Keep these figures as calibration lines in planning sheets, discussed from the Apple Ads hub, never as targets your account must hit.

Category splits change the maths#

Category medians diverge widely. US medians in estimated vendor data print near 4.21 dollars for Games, 3.55 for Finance, 2.05 for Shopping, 1.59 for Health, 2.06 for Education and 1.55 for Productivity, per the AppTweak benchmarks. A finance app bidding to a shopping CPT will starve, while a productivity app bidding to a games CPT will overpay for months.

Set opening bids and CPA caps from your own category row, then adjust to your payer mix. Our max CPT guide and CPA cap guide turn the row into working bids.

US versus UK reads#

The US runs deeper auctions and higher CPTs than the UK in most categories, with richer keyword depth and faster fatigue. The UK tracks the US pattern at slightly lower CPTs, with spelling and slang differences that reward separate lists. Keep the two markets in separate campaign groups with separate caps, scoped via location targeting.

Judge each market against its own four week history first and vendor medians second. A UK account beating its own trend while trailing a US median is winning, not losing.

Keep keyword builds separate even where terms look identical, because tap through and conversion behaviour diverge by market and blended lists blur the read within weeks.

Rank and revenue scope notes#

Two locked scope notes keep planning honest. AppsFlyer 2025 ranks Apple Ads number one on iOS for gaming and non gaming, per the AppsFlyer Performance Index, which speaks to signal quality on Apple hardware rather than to any cost promise. And the eMarketer US ad revenue series covers all Apple ads products combined, so never read it as an Apple Ads revenue figure, since Apple discloses no official Apple Ads revenue, advertiser count or market share.

Hold those boundaries in every deck and forecast. Our reconciliation guide keeps vendor, store and MMP rows apart, and our null rates guide explains the gaps that remain.

Turn medians into working targets#

Convert medians into working bids in three steps: pick your category row, haircut or premium it by your payer value, then cap discovery while exact proves out. Review fortnightly against trial starts and paid mix, moving bids in small steps that preserve learning. Our budget maths sizes the test behind the bids.

For illustration, a shopping app might open near its 2.05 dollar category median, then ease bids down as exact harvests prove payer value over a full month. Treat that framing as illustrative method, and ask our team for a benchmark review.

Frequently asked questions#