How much do Facebook app install ads cost in 2026?

Cost per install (CPI) on Meta in 2026 ranges from $0.80 to $3.50 for iOS and $0.30 to $1.50 for Android, depending on geo, category, and creative quality. Gaming apps tend to sit at the lower end, while fintech and health apps pay more. Optimizing for downstream events (purchase, subscription) raises the effective CPI but improves unit economics.

Should I optimize for installs or in-app events?

Start with App Install optimization if your app is new and the pixel has fewer than 50 weekly events. Once you accumulate 50+ weekly purchase or subscription events, switch to App Event Optimization (AEO) or Value Optimization (VO). AEO targets users likely to complete a specific in-app action, while VO targets users likely to generate the highest revenue.

Do I need the Facebook SDK for app install campaigns?

Yes. The Meta SDK is required for app install campaigns. It tracks installs, passes in-app events back to Meta for optimization, and enables app event audiences for retargeting. Without it, Meta cannot optimize beyond the click and you are effectively running blind. For iOS, you also need to configure SKAdNetwork and Apple's ATT prompt.

How does iOS 14.5+ affect Facebook app install campaigns in 2026?

Apple's App Tracking Transparency framework limits deterministic user matching on iOS. In 2026, roughly 25-35% of iOS users opt in to tracking. Meta uses Aggregated Event Measurement and modeled conversions to fill the gap. Practical impact: iOS CPI runs 40-60% higher than Android, attribution windows are shorter (1-day click by default), and you lose granular breakdown data. Most app advertisers run separate iOS and Android campaigns to manage this.

Field Notes Updated May 2026 10 min read

Facebook Ads for Mobile App Installs: The Real Playbook in 2026

App campaigns have their own attribution challenges, weird quality dynamics, and a specific creative format that crushes everything else. What works after iOS 14 changes settled.

Mobile app install (MAI) campaigns on Facebook look similar to other campaigns in setup, but the post-iOS 14 attribution mess and the quality-over-volume math make them their own beast.

The post-iOS 14 reality

SKAdNetwork (SKAN) 4.0 is the framework Apple gave us. Compared to the pre-iOS 14 deterministic pixel:

Conversions are aggregated and delayed (24-72 hour postbacks)
You can't do user-level attribution for iOS users without ATT consent (which most users decline)
Data is available at campaign level only for SKAN, not ad set or ad
Conversion values are coarse (one of a small set of buckets)

Net effect: you can still drive performance, but optimization is blunter. Test bigger swings less often, accept measurement noise.

Campaign setup that works

Use App Promotion campaign objective (not the older "App Installs")
Optimize for in-app event (Purchase, Subscribe, Level Reached) not just install — quality matters more than volume
Separate campaigns by OS (iOS uses SKAN, Android still has fairly accurate tracking)
Set up Aggregated Event Measurement priority correctly for iOS
Implement SDK or use MMP (AppsFlyer, Adjust, Singular) for proper attribution

Creative for app installs

The creative format that consistently wins: 9-15 second vertical video showing actual app gameplay/usage. Not a polished commercial. Phone-shot screen recording with brief overlay text.

Why: users are deciding "do I want to install and use this for the next month?" They need to see what using it actually looks like. Polished commercials hide the user experience.

Hook in first 1-2 seconds: show the most interesting/unique moment of the app. Don't open with brand. Open with the wow.

Real numbers from a utility app. iOS users cost 2x as much to acquire but spend 2.3x more — better unit economics despite higher CPI.

Optimization events

Don't optimize for Install. The bidder will find cheap installs from low-quality users who never engage. Optimize for the event that proves quality:

Day 1 retention event (Open Day 2)
Tutorial completion
First level / first task
First purchase / subscription

Need 50+ events/week for the bidder to optimize properly. If you don't have that volume, optimize for an upper-funnel event with more volume.

What kills app install accounts

Optimizing for Install instead of quality event — drives garbage installs
Same campaign for iOS + Android — cleaner separated
Polished video creative — performs worse than authentic gameplay
Ignoring SKAN setup — leaves iOS optimization blind
No MMP — can't measure true LTV by source

FAQ

Should I use Advantage Shopping campaigns for apps?

For shopping apps with strong catalogues, yes. For most other app categories, App Promotion gives you cleaner control.

How long do iOS conversions take to report?

SKAN postbacks come 24-72 hours after the conversion event. Plan for 3-day delay in iOS data.

Why is my CPI suddenly higher?

Usually creative fatigue (rotate creatives weekly), audience saturation, or seasonal CPM spikes.

Bottom line

App install is a different game than web conversion. Optimize for quality, not volume. Use authentic in-app footage as creative. Separate iOS and Android. Accept that measurement is bluntier than it used to be — make decisions on weekly trends, not daily noise.