We run close to $100K a month on Apple Search Ads. Here are 12 things nobody tells you.

Apple Search Ads is the most honest ad platform in mobile: the person searching has already decided they want an app, and you are bidding to be the one they see. It is also full of small traps that only show up once you spend real money. These are the twelve we would want someone to have told us.

1. Your brand campaign is lying to you, politely

Installs from people searching your own app's name would mostly have happened anyway. If brand terms share a campaign with everything else, the blended cost per install looks wonderful and tells you nothing. Separate it on day one.

2. Search Match is a research tool

Letting Apple match your ad to searches automatically is useful for finding terms you did not think of. It is a terrible way to spend most of a budget. Run it small, harvest what converts into exact match, and add the rest as negatives.

3. "Broad match" is broader than you think

A broad match keyword will happily show your bedtime story app to someone searching for a sleep tracker. Check the search terms report weekly. The first time you do it you will find money leaking somewhere surprising.

4. One CPA target for the whole account is a mistake

An install in one country can be worth several times an install in another to a subscription app, and the auction prices differ just as much. Set targets per storefront from your own revenue data, or you will quietly overpay in some markets and starve others.

5. Installs are the wrong number

Installs are what the dashboard shows by default, which is exactly why they are dangerous. Connect attribution so you see trial starts, paid conversions and first renewals by keyword. The keywords with the cheapest installs are often not the ones with the cheapest customers.

6. Your competitors' names are expensive and worth it, sometimes

Bidding on a competitor's name costs more per tap and converts worse than category terms. It still earns its place when your product genuinely answers the same need better. Test it in its own campaign, judge it on paid conversions, and keep it or cut it on that basis alone.

7. The product page decides your cost per install

Two apps with identical bids and keywords can have cost per install that differs by a wide margin, purely because one store page converts and the other does not. Before you raise a bid, look at your first screenshot.

8. Custom product pages are the biggest lever most accounts never pull

Someone searching "stories in my own voice" should land on a page whose first screenshot is about exactly that. Build a product page per intent and point ad groups at them. The conversion difference is usually larger than anything bids can achieve.

9. Negatives are half the work

A healthy account has a long, boring negative keyword list that grows every week. If yours has ten entries, the account is not being managed; it is being watched.

10. Impression share tells you when to stop

If you already win most of the impressions on a term, more bid buys very little. Check impression share before raising bids on the keywords that matter, and put the extra budget into terms where you are barely visible.

11. Weekends and launches move the auction

Competitor launches, category features and seasonal demand change prices without warning. A bid that was efficient three weeks ago may not be today. Review on a fixed weekly rhythm, not when something feels wrong.

12. The account should be yours

If an agency runs campaigns inside its own account, the keyword lists, the history and the learning leave with them. Insist that everything runs in your Apple Search Ads account with the agency as a manager. We work this way with every partner, and we would not accept anything else for our own apps.

If you want someone to go through your account with this list in hand, here is how we work. The audit is free and written, and it is useful whether or not we end up running the account together.