See every keyword from impression to revenue, then cut what doesn't pay and scale
what does. Do it yourself, or let your agent do it inside limits you set.
you › Find the Apple Ads spend that isn't paying off.→ list_recommendations
Pause "weather radar" $212 spent · 0 subscribers
Raise bid "tide chart" D30 ROAS 1.42x
→ preview_action
pause "weather radar" · within your limits
→ commit_action
applied · logged as agent "codex"
Stop wasting spend
A short, ranked list of keywords to pause, bids to cut, and search terms to block.
Impression to revenue
Every keyword followed through taps, installs, trials and paid subscribers to real proceeds.
For you and your agents
The same numbers and decisions in the dashboard and over MCP, for any agent.
Safe, auditable actions
Every change is previewed, kept within your limits, and logged with its evidence.
Demo data.
Stop wasting spend
Cheap installs that never subscribe are the most expensive ones.
SpendDaddy ranks what to change, most important first, with one sentence on why and
the numbers behind it.
Pause a keywordthat keeps spending without paying subscribers.
Block a search termthat spends and never installs.
Lower a bid or budgetwhen ROAS stays below your target.
Raise a bid or budgetwhen a winner is held back.
Promote a search termthat converts into its own exact-match keyword.
Not enough data yet? It tells you to keep watching instead of guessing.
Demo data.
Impression to revenue
Apple Ads stops counting at the download. SpendDaddy doesn't.
Spend from Apple Ads and revenue from RevenueCat, joined per keyword. Watch each
install cohort earn over 7, 30, 60 and 90 days, down to the ad group and search term.
Missing data shows as unknown with the reason, never as zero, and never drives a cut.
tide chart Last 90 days · demo data
Impressions18,420
Taps1,1526.3%
Installs43137%
Trials11827%
Paid subscribers3731%
30-day proceeds$586.20
Spend $412.80 · ROAS 1.42x · cohorts mature
For you and your agents
Your agent does the work. You set the limits.
Connect Claude, Codex or any MCP client. It sees the same numbers and evidence you
do, previews a change, applies it only if you allowed it, and checks back later.
Keys are read-only unless you grant changes, per app, with an expiry date.
You pick the campaigns, the kinds of change, maximum bids and changes per day.
Agents make small changes only. Anything bigger goes to a person.
Every change is logged with who made it and the evidence they saw.
SpendDaddy only reads your data until you turn changes on. Each change is previewed,
confirmed and logged. After a bid or budget raise, SpendDaddy watches the extra spend
and messages you on Telegram to revert or keep it if a limit is crossed.
Credentials are stored encrypted.
Can an agent overspend?
An agent can only change one keyword's bid or status, or one campaign's budget or
status, on the campaigns you pick and within the limits you set. Anything else is
refused. Limits cap configured bids and budgets; Apple still decides actual spend
within them.
Do I need an SDK or an attribution provider?
No. SpendDaddy works from Apple Ads and RevenueCat directly. An optional
server-side call at install time makes cohorts more accurate.
Why it helps
What does “revenue” mean here?
Proceeds after Apple's commission, earned in the first 30 days after install.
Cohort views also show 7, 60 and 90 days.
How long until I see recommendations?
The first import usually takes a few minutes and brings in up to 90 days of
RevenueCat revenue per keyword. We email you when your numbers are ready.
Recommendations need enough spend and mature installs, so expect them to firm up
over the first few weeks.
Who is it for, and what does it cost?
Indie developers and small studios running Apple Ads and selling subscriptions
through RevenueCat. It's free during the beta.
