7 min read ·
Apple Search Ads for Indie Developers: Start Small
Quick answer
Set a small monthly budget, start on exact match and work out the tap price you can afford before you bid. A worked break-even example for indie iOS apps.
Most indie developers come to Apple Search Ads with one question: is it worth spending money on, for an app this small? The honest answer depends on a number you can work out before you spend anything, which is the most you can pay for a tap and still get your money back. This guide starts there, then sets a budget, a first campaign structure and a weekly routine you can keep up on your own.
Apple renamed the product Apple Ads in 2025, and most developers still search for it by the old name. The ads that matter here are the ones at the top of App Store search results, paid per tap.
Work out your break-even tap price first
Every decision later in this guide leans on one figure: the highest cost per tap (CPT) at which a tap still pays for itself. Three numbers from your own consoles give it to you.
- Net revenue per paying user. What a payer brings you in the first year, after Apple's cut. If you are in the App Store Small Business Program, Apple keeps 15%, so a $29.99 yearly subscription leaves you about $25.49.
- Install-to-paid rate. The share of new installs that start paying within, say, 30 days. RevenueCat shows this as a conversion figure; App Store Connect gets you close with installs and first-time subscriptions over the same window.
- Tap-to-install rate. The share of ad taps that become installs. Apple Ads reports it per campaign and per keyword once you are running. Before you start, use your product page conversion rate from App Store Connect as a stand-in, and replace it after the first fortnight.
Multiply them:
break-even CPT = net revenue per payer × install-to-paid rate × tap-to-install rate
A worked example, with numbers made up for the arithmetic (put your own in):
| Input | Example value |
|---|---|
| Net first-year revenue per payer | $25.49 |
| Install-to-paid rate | 4% |
| Revenue per install | $25.49 × 0.04 = $1.02 |
| Tap-to-install rate | 50% |
| Break-even CPT | $1.02 × 0.50 = $0.51 |
So this app can pay up to about 51 cents a tap and break even on first-year revenue. It counts nothing after year one and nothing a user brings by word of mouth, which makes it a conservative ceiling. Keep it conservative while you learn. You can raise it once renewals show up in your own data rather than in a guess.
If your break-even CPT comes out far below what keywords in your category cost, the usual problem is upstream of the ads: a weak paywall, a trial nobody converts from, a price that is too low for paid acquisition. Fix that first. No bid strategy rescues a revenue per install of 20 cents.
Write the figure down. You will compare every keyword against it.
Set a monthly budget you are willing to learn with
Pick a monthly figure you could lose entirely without minding much, and think of it in taps rather than dollars. At a $0.80 average CPT, $300 a month buys roughly 375 taps. That sounds like a lot until you spread it over twenty keywords, at which point each keyword gets under twenty taps a month, and twenty taps tells you very little.
So the budget decides how many keywords you can test. A small budget means few keywords, chosen carefully. Divide the monthly figure by 30 for a daily budget per campaign, and leave a margin so one good day does not empty the month.
Start on exact match, with four small campaigns
A common structure, and the one we lay out when a maker asks Kwern to start their campaigns, splits keywords by intent:
- Brand. Your app's name and its misspellings, on exact match. Cheap, high conversion, and it stops a competitor bidding on your name from taking the top slot above your own listing.
- Category. The three to ten phrases that describe what your app does, the same ones that should already sit in your title, subtitle or keyword field. Exact match.
- Competitor. The names of apps yours replaces, on exact match. These tap often and convert less, so they get the smallest bids.
- Discovery. Broad match and Search Match at a low bid. Its job is to find the search terms you had not thought of, and it is the one campaign allowed to spend on terms you did not choose.
Exact match first, because a small budget cannot afford to pay for taps on searches you would never have chosen. Broad match buys reach. You can buy reach later, once you know which terms earn their tap.
Two settings save money on day one. Add every exact keyword from the other three campaigns as a negative keyword in Discovery, so you do not bid against yourself. And set each keyword's starting bid at or just under your break-even CPT, even when Apple's suggested bid is higher. A suggested bid tells you what others are paying. Your break-even says what you can afford.
If you are unsure what belongs in Category, our guide to filling the App Store keyword field covers how to pick the phrases your listing should rank for. Category campaigns and the keyword field draw on the same list.
Read it once a week, against revenue
Apple Ads changes are slow to show what they did, and a bid judged after two days is judged on noise. A weekly read, at the same time each week, is enough for an account this size. Each week:
- Open the search terms report. In Discovery, look for terms that took taps and installed. Move each one into Category or Competitor as an exact keyword, and add it as a negative in Discovery. Look for terms that took taps and installed nobody, and add those as negatives too.
- Compare cost per install to revenue per install, per keyword. Cost per install comes from Apple Ads. Revenue per install comes from RevenueCat, whose Apple Search Ads attribution ties a subscriber back to the campaign and keyword that brought them. This is the comparison that matters. A keyword with a high CPT and paying users beats a cheap keyword whose installs never open the paywall.
- Make one change per keyword, at most. Lower the bid on a keyword whose cost per install is above its revenue per install. Raise it a little on one that is under break-even and losing impressions. Keep each move small, about a fifth of the current bid, so you can tell what the move did.
- Pause what has proven nothing. A keyword that has taken three weeks of steady taps with no installs is telling you something. Pause it and write down what it spent.
- Check the month. Spend so far against the monthly budget. If a campaign runs out of daily budget early on days it converts well, that is the one to give more.
Keep a short log: the date, the keyword, the old bid, the new bid and the reason. Two months in, that log is the only reliable record of what worked, because the console shows you today's numbers and nothing about why they changed.
Know when to stop
Give it six to eight weeks at a budget large enough for your main keywords to collect meaningful taps. After that, look at blended cost per install against blended revenue per install. If the second is clearly above the first, raise the budget slowly and add keywords from Discovery. If it is below and the gap has not narrowed, stop or cut to Brand only. Brand alone often pays for itself, and spending on it means your own listing stays above a competitor's ad.
Stopping also tells you something. An app that cannot pay for installs at its category's tap prices has a monetisation question to answer, and that is worth more of your time than another month of bids.
Where Kwern fits
You can run everything above with a spreadsheet and thirty minutes a week, and plenty of developers do. The part that slips is the weekly read: the account gets set up in launch week and is not opened again.
Kwern's Apple Ads agent does that read for you. Once you connect Apple Ads, it reads your campaigns, keywords and search terms report each week, sets them against the monthly budget you gave it, and proposes the moves this guide describes: a lower bid, a higher budget, a search term added as an exact keyword, a keyword paused. Each proposal carries the old value, the new value and the reason, and nothing changes in your account until you approve it. With RevenueCat connected, it judges a keyword on the revenue behind the installs rather than on taps alone. Campaigns themselves start in a chat, where Kwern lays out Brand, Category, Competitor and Discovery as proposals you publish one by one.
It is off until you turn it on, because it can move money. Plans start at $29 a month, with the model on your own provider key; the pricing page has the rest.
Whichever way you run it, work out the break-even tap price before the first dollar goes out. Every bid afterwards is a comparison against that number.