Apple Search Ads (ASA) help promote your iOS app directly within the App Store’s search results. With iOS privacy updates (especially iOS 14.5+), campaign optimization relies on SKAdNetwork (SKAN), which limits granular attribution but ensures user privacy. Here’s how ASA works and how to optimize campaigns using SKAN:

---

🔍 How Apple Search Ads Work

1. Placement: Your ad appears when a user searches for a relevant keyword in the App Store.

2. Matching:

   • Exact Match: Triggers ad only for that specific keyword.

   • Broad Match: Covers similar and related search terms.

   • Search Match: Apple auto-matches to relevant queries.

3. Auction-Based: CPC (Cost Per Tap) model; you bid on keywords.

4. Attribution:

   • Apple Search Ads measures installs directly on iOS.

   • For non-consenting users (ATT opt-out), installs are measured via SKAN.

   • Campaigns are mapped to SKAN campaign IDs (limited to 100 per ad account).

---

📈 How SKAdNetwork (SKAN) Works in ASA Context

• SKAN Attribution:

  • Apple sends postbacks with anonymized install data after a 24–72 hour window.

  • Postbacks contain limited data: conversion value (0–63), campaign ID, and app ID.

  • No real-time user-level data.

---

⚙️ How to Optimize ASA Campaigns via SKAN

✅ 1. Structure for SKAN

• Limit active campaigns to under 100 (SKAN campaign ID limit).

• Use ad group-level granularity to test keywords within a SKAN campaign.

🧠 2. Use Conversion Values Strategically

• Map SKAN conversion values to events like:

  • Day 1 retention

  • Registration

  • Deposit

  • Tutorial completion

• Use a 64-value model smartly via MMP (e.g., Branch, Adjust) to infer user quality.

🔍 3. Split Campaign Types

• Run these separately for clarity:

  • Brand Campaigns (own app name)

  • Generic Campaigns (e.g., “rummy cash game”)

  • Competitor Campaigns (e.g., “RummyCircle”)

  • Discovery Campaigns (Search Match enabled)

🧪 4. Run SKAN + ASA Native Reporting in Parallel

• ASA provides both:

  • SKAN Attribution (privacy-safe, aggregated)

  • Apple’s Native Reporting (for opted-in users with ATT)

🧠 5. Use MMP’s SKAN Reporting Tools

• MMPs like Branch.io can:

  • Decode conversion values

  • Group users by LTV tier or funnel progress

  • Provide modeled performance reports for SKAN campaigns

📊 6. Benchmark Using ID-Optional Users

• Use ATT opt-in data to model performance benchmarks.

• Estimate SKAN ROI using opt-in data for campaigns running in parallel.

🔄 7. Rotate Creative and Keyword Testing

• SKAN has limited feedback delay; run tests for at least 7 days.

• Use multiple ad variations per ad group.

---

🚀 Advanced Tips

• ✅ Use CPA Goals where SKAN ROAS isn’t available.

• 🧪 Test custom product pages to match keywords with specific landing experiences.

• 💡 Allocate budget to SKAN campaigns that show strong early signals (e.g., high conversion value scores).

• ⏱ SKAN postbacks can delay optimization; give enough runway (~2 weeks) before making decisions.

What is SKAdNetwork (SKAN)?

Apple’s privacy framework introduced in iOS 14.5+ to allow advertisers to track installs and post-install actions anonymously, without tracking the user.

Since user-level tracking is restricted without ATT (App Tracking Transparency) consent, SKAN gives you aggregated campaign data through a single postback per user — and that postback can only carry:

• ✅ Campaign ID

• ✅ App ID

• ✅ Conversion Value (between 0 and 63)

• ✅ Timestamp (with delay)

---

🔢 What is the Conversion Value (0–63)?

SKAN allows only one 6-bit number (0–63) to be sent once per user after install. This number is called the conversion value.

You can assign meaning to each number using your MMP (like Branch or Adjust). This allows you to represent user behavior (registrations, deposits, gameplay, etc.) using a simple number.

---

🔧 Example: Mapping Conversion Values

Let’s say you define:

You configure this logic inside your MMP dashboard (Adjust or Branch).

---

🕒 When is Conversion Value Sent?

• Apple waits 24 to 72 hours after the install before sending the conversion value.

• You can extend the timer by updating the conversion value as the user performs more actions (example: app open → registration → deposit).

• After no more updates within the 24h window, Apple finalizes and sends the postback.

---

✅ What MMPs Like Branch or Adjust Do

🧠 They:

1. Help you map conversion values to real events (in the dashboard).

2. Track user behavior in the app and update the conversion value before Apple locks it.

3. Decode the SKAN data received and aggregate performance (e.g., how many users deposited ₹500+).

4. Allow you to create SKAN LTV models for ROI-based optimization.

---

💡 Visual Analogy

Think of it like this:

• SKAN only lets you send one sealed envelope per user.

• That envelope can only contain a number from 0–63.

• You choose what each number means.

• Apple collects the envelopes (delayed), and Branch/Adjust help you read and interpret them.

---

🔄 Why It’s Important

• You can’t track users, but you can still:

  • Optimize campaigns

  • See which campaigns drive quality installs

  • Measure early indicators of LTV

• Helps you make decisions without violating Apple’s privacy rules.

---

📘 Summary