What Apple Search Ads is good for

Someone typing your category into the App Store search bar has already decided they want this kind of app. That is a different person from someone scrolling a feed, and it shows up in every downstream number: trial start rates and trial-to-paid rates from search traffic are routinely the best in the account.

It also has a property no other iOS channel has. Apple attributes its own installs deterministically through the Ads Attribution API, so you get real keyword-level data rather than SKAdNetwork's aggregated, delayed, privacy-limited picture. That makes ASA the best available place to learn things you can then apply everywhere else.

The structure we use

Four campaign types, kept strictly separate, because blending them makes every number meaningless:

Brand

Your own app name and misspellings. Cheap, high-converting, and largely defensive – you are stopping competitors from appearing above your own listing. Reported on its own so it never flatters the blended cost per acquisition.

Category and generic

The terms describing what the app does. This is where real incremental volume lives and where most of the budget belongs once discovery has done its job.

Competitor

Rival app names. Expensive per tap, lower conversion, but valuable intelligence about who you actually lose to. Worth running deliberately and small unless the cohort data says otherwise.

Discovery

Broad and search-match campaigns whose job is not conversion but to surface the terms real people type. Winners get promoted to exact match in the campaigns above; everything else gets added as a negative. This loop is the engine of the account.

What each campaign type is for
CampaignMatch typeJobReport it against
BrandExactDefend the listingCost, separately from everything else
Category / genericExact, promoted from discoveryIncremental volumeCost per trial and payback
CompetitorExactTake share, learn positioningTrial-to-paid, not install cost
DiscoveryBroad + search matchFind the real vocabularyTerms surfaced, not CPA

What usually goes wrong

Brand and generic in one campaign. Cheap brand installs drag the average down until the account looks efficient and nobody notices that generic is unprofitable.

No discovery loop. The keyword list is whatever someone thought of at setup, a year ago, and never learned anything since.

Optimising to cost per install. ASA gives you deterministic keyword-level data – the one channel where you can see which specific search term produced a paying subscriber. Stopping at install throws that away.

Ignoring the product page. You pay per tap, so conversion rate on the page decides your effective install cost. Custom Product Pages let you match the page to the search intent, and most accounts run none.

Questions people actually ask

What does Apple Search Ads cost?
You bid a cost per tap and pay per tap, not per install. What that becomes as a cost per install depends on your tap-through and conversion rates, which are driven by your product page. Branded keywords are usually the cheapest inventory you will ever buy; competitor and generic terms cost several times more.
Should I bid on my own brand name?
Almost always yes, and the reason is defensive. If a competitor bids on your name and you do not, they appear above your own listing when someone searches for you specifically. Brand traffic is also your cheapest, which flatters blended numbers – so we report it separately from the start.
Basic or Advanced?
Advanced, unless you are spending under about $2,000 a month and nobody has time to manage it. Basic hands bidding and keyword choice to Apple and caps you at $10,000 per app per month, with no keyword-level reporting – which means no way to learn anything you can reuse.
How does Apple Search Ads interact with SKAdNetwork?
It does not need it. Apple attributes its own installs deterministically through the Ads Attribution API, so ASA data is more reliable than any other iOS channel. That is exactly why it should be reported separately rather than blended into a SKAN-based view.
Which ad placements are worth using?
Search results first – highest intent, clearest measurement. Today tab and Search tab are awareness placements with much looser targeting; they can work at scale for a known brand and usually waste money for everyone else. Product page ads sit in between.
How long before it is optimised?
Discovery campaigns need two to three weeks to surface the search terms people actually use, which is the real deliverable of the first month. Structure settles by week four. After that it is steady keyword and bid work rather than dramatic change.

Free

Get your Apple Search Ads audit

What are you after

One of these. Whichever you pick, no call is required to get an answer.

What should we look at

Pick one or more. Each audit is done by hand, not by a tool.

We cannot audit anything without it.

Genuinely useful to us – half our traffic is impossible to attribute otherwise.

Free · one reply within a working day · no call needed

Related reading.

The subscription benchmarks show what the downstream numbers should look like once the traffic is good, and the paywall guide covers what to do when they are not.