How tap-through diagnoses ad fit

TTR defined, precisely#

Tap-through rate divides taps by the ad views that earned them. It sits upstream of conversion rate: TTR judges whether the ad deserved the tap, conversion rate judges whether the product page deserved the install. Ailing CPA with healthy conversion almost always starts as ailing TTR.

Keep the vocabulary straight with our TTR and CPT glossary, place TTR in the full chain through CPT to CPA maths, and orient the account from the Apple Ads hub.

Yardsticks worth trusting#

SplitMetrics 2025 panel data (estimated vendor data) reports search tap through near 9.7 per cent with conversion near 66.2 per cent, and their Apple Search Ads cost study shows the panel behind both figures. Treat that as a market sketch, then build your own norm: brand terms should beat it comfortably, generic discovery may sit below it honestly.

Your account norm beats any panel. Split TTR by tier and match type before comparing anything, because blended TTR hides the same leaks blended CPA does. Discovery versus brand sets the split, and reconciliation keeps the install side of the comparison honest.

Revisit the norm quarterly: seasonality, new rivals and listing changes all move the baseline, and a norm frozen in spring misjudges autumn traffic. Dated norms punish good keywords for market weather.

Low TTR with high views: mismatch#

Plenty of views with few taps means searchers see your ad and pass: a keyword-to-ad mismatch, not a bid shortage. Raising bids here buys more rejections at a higher price, so reach for fit tools instead.

Cut the queries that never deserved you with negatives, tighten the match with exact, and check the underlying fit through relevancy. Apple own keyword bid guidance assumes relevance before recommending raises, which is exactly this discipline.

Low TTR with low views: bid or popularity#

Few views with few taps is a different patient: the keyword may simply lack volume, or the bid may sit too low to test fit fairly. Check search popularity first; a term nobody searches needs no diagnosis, only replacement.

For popular terms that cannot win views, raise per max CPT ceilings, confirm demand in search popularity, and review coverage with impression share before spending another pound on the verdict.

TTR behaves differently by match type#

Exact match should post the highest TTR, broad match a middling one, and Search Match the lowest, because each step outward trades precision for reach. Grade each match against its own band: broad TTR that beats your exact norm is a promotion candidate, not a problem.

Learn the bands in broad match and Search Match, promote proven queries upward through harvesting, and price each band through bid tiers.

The weekly TTR repair routine#

Once a week, sort keywords by views, flag high-view low-TTR terms, and act: negate, tighten, or rehouse each one. Ten focused minutes beats ten scattered tweaks, because the worst offenders dominate the waste.

Localise the repair where queries cross languages via localised keywords, mind the keyword count ceiling with limits, and if the flags never clear, our team runs TTR triage as a standing weekly service.

Close the loop the following week: re-pull the flagged terms and confirm the fix moved the rate before declaring victory. Repairs without verification are wishes with timestamps.

Frequently asked questions#