Setting a sane starting max CPT bid for a new ASA campaign

You're about to launch a new Apple Search Ads campaign. The bid field is blinking at you. Type 50 cents and you might never get a tap. Type five dollars and you might empty your daily budget on three irrelevant searches. So what number actually goes in there?

There's no universal right answer, but there is a sane way to think about it. The goal of your first bid isn't to win — it's to collect enough data to make a smarter bid next week. Here's how I approach it.

Start from economics, not from gut feel

Your max CPT should be anchored to what a tap is actually worth to you. Work backwards from revenue, not forwards from "what feels reasonable."

The rough chain is:

Tap → install: your tap-through-install rate (conversion rate on ASA).
Install → paying user: your free-to-paid conversion in the app.
Paying user → revenue: average revenue per paying user over whatever window you care about (30 days, 90 days, or LTV).

Multiply those together and you get revenue per tap. That's your absolute ceiling. Your max CPT has to live well below it if you want this campaign to be profitable.

A worked illustrative example (your numbers will differ):

Conversion rate on ASA: 50% of taps install.
Of those installs, 3% become paying within 60 days.
Average 60-day revenue per payer: $40.
Revenue per tap = 0.50 × 0.03 × $40 = $0.60.

If your revenue per tap is around 60 cents, bidding $1.50 max CPT is lighting money on fire. Bidding 10 cents probably means you never enter the auction. Somewhere in the 20–40 cent range is a defensible starting place — leaving margin for the fact that early data is noisy and you'll lose some users who never come back.

If you don't have your own conversion data yet, use conservative placeholder assumptions and accept that your first week is paid research. Set a small daily cap so the research bill stays small.

Treat your max CPT as a ceiling, not a price

A common misconception: the max CPT is what you'll pay. It isn't. Apple runs a second-price-style auction, so you usually pay less than your max. Your max is the most you're willing to pay for a tap on that keyword.

That means:

Bidding higher doesn't necessarily mean overpaying on every tap — it means you're eligible for more auctions.
Bidding too low doesn't save you money; it just means you don't show up.
The goal is to set a ceiling that reflects the keyword's value to you, then let the auction settle below it.

Different keywords deserve different bids

A single flat bid across an ad group is fine for week one, but it's almost never optimal. Intent varies wildly:

Brand keywords (your app name, competitor names): highest intent. Searchers who type your competitor's name and convert are worth a lot. These can support the highest bids.
Category keywords ("habit tracker", "budget app"): medium intent, broader audience, usually lower conversion rates but more volume. Bid lower.
Generic/discovery terms ("productivity", "finance"): low intent, often poor conversion. Bid lowest, or skip entirely until you've proven the rest.

A reasonable starting structure: group keywords by intent tier and give each tier its own ad group with its own bid. Then you only have to tune three numbers, not thirty.

Set up the campaign so the data is actually usable

Bids only get smarter if you can read what happened. Before you launch:

One country per campaign. This is required anyway, but it also means you can't accidentally compare a US tap to an Indonesian tap at the same bid.
Separate Search Match into its own ad group with its own (usually lower) bid. Search Match will find queries you didn't think of, and you want to see them isolated, not mixed in with your exact-match keywords.
Use exact match for keywords where you have a strong hypothesis. Use broad sparingly at first — it expands reach but muddies attribution to specific search terms.
Wire up install → revenue attribution (via AdServices and a tool like RevenueCat) before spending real money. Without that, you can only optimize to installs, and installs lie.

A reasonable starting heuristic

If you're truly starting from zero, here's a rough recipe — not a rule, just a place to begin:

Estimate revenue per tap from the formula above. If you really have no data, pick a deliberately pessimistic number.
Start your max CPT at roughly 40–60% of estimated revenue per tap for category keywords. This gives the auction room while keeping you under water-line on a per-tap basis.
Bid higher (70–100% of estimated revenue per tap) on brand/competitor keywords where intent is strongest.
Bid lower (20–30%) on Search Match and generic discovery terms.
Set a daily cap that, if completely spent, you'd be fine losing as research budget for one to two weeks.

Then don't touch it for several days. Bid changes need a few hundred taps to mean anything; reacting after a day of data is how you trick yourself into chasing noise.

When and how to adjust

After you've got a meaningful sample (this depends on your scale, but think hundreds of taps per keyword, not dozens), look at each keyword on three axes:

Impressions: are you even showing up? If no, bid is probably too low or the keyword has no volume.
TTR: if you're getting impressions but no taps, the issue is your app's appearance in results (icon, name, ratings) — not your bid.
Conversion rate and CPA: if taps are cheap but no one installs or pays, lower the bid or pause. If CPA is well under your target, raise the bid carefully (10–20% at a time) to capture more volume.

Resist the urge to make ten changes at once. One variable per adjustment, then wait.

A note on tooling

Manually re-running this analysis across dozens of keywords every week is tedious, and it's the kind of thing that quietly slips when you're shipping features. This is roughly what we built AdsBuddy to handle — it reads your ASA spend alongside your actual revenue and surfaces a short list of suggested bid changes each day, with the reasoning. You still approve every change yourself; nothing moves without your say-so. But the bid math above is exactly the math it's doing in the background, on every keyword, every day.

Takeaway

Your first max CPT isn't a commitment — it's a hypothesis. Anchor it to revenue per tap, tier it by keyword intent, structure the campaign so the data comes back clean, and give it enough time to mean something before you change it. Done that way, your week-two bids will be dramatically better than your week-one bids, and you'll keep compounding from there.