Ad Budget Calculator for App Installs

App installs are the only paid-media goal where the platforms hide your data on purpose. SKAdNetwork postbacks, modeled conversions in UAC, and 24-72 hour reporting delays mean your budget plan has to assume you're flying half-blind for the first two weeks.

The confused-goal trap on this page. A surprising number of teams pick "app installs" when they actually want in-app purchases or subscriptions. If your monetization happens 7+ days after install, optimizing campaigns toward install events will drown you in tire-kickers. Switch the optimization event to purchase, trial_start, or a strong proxy (tutorial complete, account created) the moment you have 30-50 of them per week per ad set. UAC's tCPA and Meta's value optimization both need that volume to function — below it, fall back to install-optimized campaigns and accept the noise.

Channel mix that actually works in 2025. The viable list is short: Apple Search Ads, Google App Campaigns (UAC), Meta Advantage+ App, TikTok, and AppLovin/Mintegral for gaming. Everything else (Snap, Reddit, Pinterest UA) is a rounding error unless your audience lives there specifically.

• Apple Search Ads: $2-$8 CPI, highest intent, brand-defense first then category keywords
• Google UAC: $1.50-$4 Android, $4-$10 iOS, best for scale, worst for control
• Meta Advantage+ App: $3-$8 iOS, $2-$5 Android, strong for consumer/social apps
• TikTok: $3-$12 CPI, creative-driven, mandatory for anything Gen-Z
• AppLovin/Mintegral: gaming only, $0.50-$3 CPI, but watch fraud rates

Bidding strategy by maturity. Pre-product-market-fit (under ~500 daily installs): use Max Installs / Install Volume bids and tight geo limits. You're buying signal, not LTV. Post-PMF: shift to tCPA targeting your downstream event (tCPA on purchase, tROAS in UAC and Meta). Don't move to value-based bidding until you have at least 50 conversions per week per campaign — Google and Meta both need that to model.

LTV:CPI is the only ratio that matters. Healthy ratios in 2025:

• Hyper-casual gaming: 1.2:1 to 1.5:1 (you're betting on volume)
• Mid-core gaming: 2:1 to 3:1 at D90
• Subscription consumer apps: 3:1 at D180
• Fintech/banking: 4:1+ (the install is cheap relative to a funded account worth $80-$200)

If you can't model 90-day LTV yet, use a proxy: D7 retention × ARPDAU × 90. It's not right, but it's directionally sane and it'll keep you from nuking your runway in week three.

Attribution windows are not what you think. SKAN 4 gives you a 0-2 day, 3-7 day, and 8-35 day postback window, but you have to configure conversion values manually and most teams misconfigure them. Use an MMP (Adjust, AppsFlyer, Singular) — running UA without one in 2025 is malpractice. On Android, the Play Install Referrer is cleaner but still gets nerfed by Privacy Sandbox changes. Build your dashboards around installs by source from the MMP, not platform-reported numbers, which will overstate by 15-40%.

What success looks like. Month 1: blended CPI within 30% of your target, MMP-attributed installs within 20% of platform-claimed. Month 2: at least one channel hitting tCPA on a downstream event. Month 3: clear winner channel taking 50%+ of budget, with a second channel proven for diversification. If by month 3 you're still spreading budget evenly across 4 channels, you don't have a winner — you have noise.

What you can't do at most budgets. Below $10K/month you can't run more than 2 channels well. Below $25K/month you can't sustain a creative refresh cadence fast enough for TikTok or Meta. Below $50K/month you probably shouldn't touch incrementality testing — the confidence intervals will eat your conclusions.

Example

A subscription language-learning app at $20K/month on iOS-primary: $7K Apple Search Ads (brand + 40 category keywords), $6K Meta Advantage+ App optimized for trial_start, $5K TikTok with 6 UGC creatives refreshed biweekly, $2K Google UAC iOS for incremental reach. Target: $6 blended CPI, $22 cost per trial, 35% trial-to-paid, $7.70 CAC against $42 12-month LTV.

Questions