How to Set Up Facebook App Install Campaigns (2025 Guide)

The average CPI on Facebook ranges from $1-$5 globally, but your actual cost depends almost entirely on how you set up your campaigns.

Most apps overspend in the first few weeks because they rush through setup without understanding how Facebook's optimization system works. The platform needs specific signals to deliver efficiently, and missing even one configuration step can increase your CPI by 30-50%.

Here's how to set up Facebook app install campaigns correctly from the start.

Prerequisites Before Campaign Setup

Before you create your first campaign, three foundational elements need to be in place.

1. Facebook SDK Integration

Your app needs Facebook SDK version 16.2.1 or later to support SKAdNetwork 4.0 and proper event tracking. Without this, Facebook can't optimize delivery or measure performance accurately.

For iOS apps specifically, you'll need to handle ATT (App Tracking Transparency) prompts and configure SKAdNetwork postbacks. For Android, ensure you're passing the correct app event data through the SDK.

2. Events Manager Configuration

Connect your app in Facebook Events Manager and verify that install events are firing correctly. This is where you'll map your in-app events to Facebook's optimization system.

Test your implementation by installing your app on a test device and confirming the install event appears in Events Manager within 24 hours.

3. App Store Presence

Your app needs to be live on the App Store or Google Play. Facebook requires a publicly accessible app listing to run app install campaigns.

Make sure your app store page is optimized before running paid traffic to it—your conversion rate directly impacts your CPI.

Campaign Structure Setup

Facebook's campaign structure operates on three levels: Campaign, Ad Set, and Ad. Each level controls different aspects of delivery.

Campaign Level: Objective and Budget

Create a new campaign and select "App Promotion" as your objective. This unlocks app-specific optimization options that aren't available with other objectives.

For budget settings, you have two options:

Campaign Budget Optimization (CBO): Facebook distributes your budget across ad sets automatically. Best for scaling once you've identified winning audiences.

Ad Set Budget Optimization (ABO): You control budget allocation manually. Best for testing when you're still learning which audiences perform.

Start with ABO if this is your first campaign. Once you identify winning ad sets, migrate to CBO for efficient scaling.

Budget recommendation: Start with $10-$15 per day per ad set. This gives Facebook enough volume to exit the learning phase without overspending during the testing period.

Ad Set Level: Targeting and Optimization

Create 3-5 ad sets to start. More than this spreads your budget too thin and slows down learning.

Optimization Event: For new campaigns with limited data, optimize for "App Installs." Once you accumulate 50+ conversions for a specific in-app event like purchases or registrations, switch to optimizing for that higher-value action.

Targeting: Use broad targeting when starting. Select your core demographic (age, gender, location) and leave interests blank. Facebook's algorithm identifies your best converters faster with broader targeting.

If you have existing customer data, create a 1% lookalike audience from your highest-value users and test it alongside broad targeting.

Placements: Enable automatic placements. Facebook's data shows automatic placements consistently deliver lower CPIs than manual selection because the algorithm finds the most cost-efficient inventory across all placement options.

Ad Level: Creative Assets

Upload 2-4 video or image creatives per ad set. If using video, keep it under 15 seconds—completion rates drop significantly after that.

Your creative should communicate value in the first 3 seconds. Users scroll quickly, and most never unmute video ads.

Include captions on video ads. Approximately 85% of Facebook video is watched without sound.

Use clear, direct calls-to-action. "Install Now" or "Download Free" performs better than vague CTAs like "Learn More."

Optimization Settings

These settings significantly impact performance but are often overlooked during initial setup.

Delivery Optimization

Under "Optimization & Delivery," select "App Installs" as your optimization event for new campaigns. Facebook needs at least 10 app installs within 7 days to exit the learning phase (reduced from 50 in 2024).

Once you exit learning, performance stabilizes and CPIs typically decrease by 15-25%.

Bid Strategy

Use "Lowest Cost" bidding when starting. Facebook automatically bids to get you the most installs within your budget.

Avoid setting a cost cap initially—you don't have enough data yet to know what an achievable CPI is for your app and audience.

After running for 7-14 days, review your average CPI. If it's within your target, continue with lowest cost. If it's consistently below target, consider setting a cost cap slightly above your current average to potentially increase volume while maintaining efficiency.

iOS 14.5+ Settings

Enable the iOS 14.5+ campaign toggle in your campaign settings. This allows Facebook to reach users on iOS 14.5 and higher, significantly expanding your available audience.

Without this enabled, your campaigns will only deliver to users who opted into tracking—typically 15-25% of iOS users.

Budget Formula and Expectations

A practical budget formula: (Target Daily Installs × Expected CPI × 1.5)

The 1.5 multiplier accounts for learning phase inefficiency and day-to-day fluctuation.

Example: If you want 10 installs per day and expect a $3 CPI based on industry benchmarks, set your daily budget to $45.

During the first 7 days, expect your CPI to run 30-50% higher than it will once you exit learning. This is normal—the algorithm needs data to optimize delivery.

Post-Launch Monitoring

After launching your campaigns, avoid making changes for the first 3-5 days. Every edit resets the learning phase and delays optimization.

Monitor these metrics daily:

Install volume: Are you getting enough installs to exit learning within 7 days?

CPI trend: Is your cost per install decreasing over time?

Frequency: If frequency exceeds 3.0 within the first week, your audience is too narrow.

If you're not generating at least 2-3 installs per day per ad set after 5 days, either increase budget or broaden your targeting. Low volume prevents the algorithm from learning effectively.

Current Benchmarks

Metric	Benchmark
Average CPI (Global)	$1-$5
CPI Range (72% of advertisers)	$0.01-$5.00
Gaming/Finance Apps	$5-$10+
Learning Phase Duration	7 days (10 installs)
CPI Improvement Post-Learning	15-25% decrease

Source: Business of Apps, Meta Business Help Center (2024-2025 data)

FAQs

What's a good CPI for Facebook app install campaigns?

Average Facebook app install CPI ranges from $1-$5 globally, with 72% of advertisers falling within the $0.01-$5.00 range. Gaming and finance apps typically see higher CPIs of $5-$10+ due to increased competition.

What budget do I need to start Facebook app install campaigns?

Start with minimum $10/day per ad set. Use the formula: (Target Daily Installs × Expected CPI × 1.5). For 10 installs per day at $3 CPI, budget $45/day.

Should I use automatic or manual placements?

Use automatic placements when starting. Facebook's algorithm optimizes delivery across placements based on performance, typically resulting in 15-30% lower CPIs than manual placement selection.

How long does it take to see stable performance?

Expect 7-14 days for campaigns to exit the learning phase and deliver stable CPIs. During the first week, CPIs typically run 30-50% higher than they will once optimization completes.

Should I optimize for installs or in-app events?

Optimize for installs if you're just starting or have fewer than 50 conversions for any in-app event. Once you hit 50+ conversions for a higher-value event like purchases or subscriptions, switch your optimization to that event.

---

Campaign setup determines everything that follows. Take time to configure correctly, resist the urge to make daily changes, and give Facebook's algorithm the data it needs to optimize delivery.