SKAdNetwork 5.0: What Changed for iOS Attribution

Apple keeps tightening. Here's what SKAN 5.0 actually changes.

Since iOS 14.5, Apple has systematically dismantled the tracking infrastructure that mobile advertisers depended on. ATT (App Tracking Transparency) gave users a choice, and 75-80% chose to opt out. SKAdNetwork became the only Apple-sanctioned way to measure iOS ad performance.

Each SKAN version has added capabilities while maintaining Apple's core principle: measure advertising effectiveness without tracking individual users. SKAN 5.0, released in 2025 and now widely adopted in 2026, represents the most significant evolution yet.

Here's what changed, what it means for your campaigns, and how to adjust your measurement stack.

A brief history: SKAN 1.0 through 4.0

For context on where we've been:

SKAN 1.0-2.0: Basic install attribution. You knew which ad network drove an app install, but with a random 24-48 hour delay and no post-install conversion data. Usable but barely.

SKAN 3.0: Added conversion value support -- a 6-bit value (0-63) that could encode post-install events. Media buyers could track whether an install led to a purchase, but the encoding was primitive and the delay remained.

SKAN 4.0: The biggest upgrade to date. Introduced three conversion windows (0-2 days, 3-7 days, 8-35 days), hierarchical conversion values (fine-grained for high-volume campaigns, coarse for low-volume), and "crowd anonymity" tiers that provided more data to campaigns with more installs. Also introduced web-to-app attribution.

SKAN 4.0's limitation: Despite improvements, it still had significant delays, limited conversion value granularity, and no support for re-engagement or view-through measurement beyond basic metrics.

What's new in SKAN 5.0

Re-engagement attribution

The single biggest addition. Previous SKAN versions only measured new app installs. SKAN 5.0 can now attribute re-engagement events -- when a lapsed user returns to your app after seeing an ad.

For many app businesses, re-engagement campaigns (targeting existing users who haven't opened the app in 30+ days) represent a major portion of ad spend. Prior to SKAN 5.0, these campaigns were essentially unmeasurable on iOS through Apple's framework.

What it means for media buyers: You can now measure re-engagement campaign performance through SKAN, with the same privacy protections (crowd anonymity, delayed postbacks) as install campaigns. This doesn't give you user-level data, but it does give you aggregate campaign-level re-engagement counts.

Expanded conversion value windows

SKAN 4.0 offered three fixed windows: 0-2 days, 3-7 days, and 8-35 days. SKAN 5.0 extends this with a fourth window (36-90 days) for campaigns that need longer measurement periods.

This matters for subscription apps, gaming apps with long monetization curves, and any app where user value develops over weeks, not days. The 90-day window lets you capture downstream conversions that were previously invisible to SKAN.

What it means for media buyers: Better LTV measurement through SKAN. You can encode revenue events from up to 90 days post-install into the SKAN postback, giving you a more complete picture of campaign value.

Higher-fidelity conversion values at lower volumes

SKAN 4.0's crowd anonymity system provided fine-grained (6-bit) conversion values only for high-volume campaigns. Low-volume campaigns received only coarse values (high/medium/low) or just install counts.

SKAN 5.0 lowers the volume thresholds for fine-grained data. Campaigns that previously only received coarse values can now access more granular conversion data, though the exact thresholds remain opaque (Apple doesn't publish them).

What it means for media buyers: Smaller campaigns and niche audience segments get usable conversion data. This is particularly helpful for testing new audiences or creative angles where initial volumes are low.

Improved web-to-app measurement

SKAN 4.0 introduced basic web-to-app attribution. SKAN 5.0 improves this with better handling of Safari-to-app flows, more reliable postback delivery, and support for web re-engagement.

For businesses that run web campaigns driving users to app stores, this closes a measurement gap that's been problematic since ATT launched.

Postback timing improvements

Previous SKAN versions added random delays of 24-144 hours before sending attribution postbacks. SKAN 5.0 reduces the maximum delay and introduces more predictable delivery windows, though exact timing is still randomized to prevent user-level identification.

What it means for media buyers: Faster data availability. Instead of waiting 2-6 days for SKAN postbacks, you'll typically see data within 1-3 days. Still not real-time, but a meaningful improvement for optimization cadence.

What SKAN 5.0 doesn't fix

No user-level data

SKAN remains an aggregate measurement framework. You get campaign-level and ad-group-level metrics, not user-level conversion paths. Individual-level attribution on iOS still requires user consent through ATT (which 75-80% deny).

No view-through attribution

SKAN 5.0 still only measures click-through attribution. If a user sees your ad (but doesn't click) and later installs your app, SKAN doesn't capture that. For platforms like Meta where view-through conversions are a significant portion of reported installs, this remains a blind spot.

No cross-device tracking

SKAN measures iOS activity only. If a user sees an ad on their iPhone and converts on their iPad or MacBook, the cross-device path isn't tracked through SKAN.

Conversion value encoding is still complex

The 6-bit (0-63) conversion value must be manually designed to encode the events you care about. Different revenue tiers, event combinations, and engagement milestones all need to be mapped into 64 possible values. This mapping is non-trivial and requires careful planning.

How to optimize your SKAN 5.0 strategy

Update your conversion value schema

With the new 90-day window, redesign your conversion value encoding to capture longer-term value signals. A recommended approach:

• Window 1 (0-2 days): First session engagement + first purchase
• Window 2 (3-7 days): 7-day retention + revenue tier
• Window 3 (8-35 days): 30-day retention + cumulative revenue tier
• Window 4 (36-90 days): 90-day LTV tier + subscription status

Each window should encode the most valuable signal for that time period. Prioritize signals that correlate with long-term value.

Set up re-engagement measurement

Configure your ad networks and MMPs to receive SKAN 5.0 re-engagement postbacks. Create a separate measurement framework for re-engagement campaigns:

1. Define what counts as "lapsed" (30 days no app open? 14 days?)
2. Set conversion values specific to re-engagement (return visit, re-purchase, subscription reactivation)
3. Establish baseline re-engagement rates before measuring lift from campaigns

Combine SKAN with server-side attribution

SKAN provides Apple's sanctioned measurement but has inherent limitations. Complement it with:

• Server-side tracking on your web properties for web-based conversions
• MMP probabilistic matching for the conversions SKAN can't capture
• Backend revenue data to validate SKAN-reported conversion values

The most accurate iOS measurement stack in 2026 uses SKAN for what it's good at (install and re-engagement attribution) and fills the gaps with server-side data.

Adapt your optimization cadence

Even with faster postbacks, SKAN data arrives 1-3 days after the conversion event. Adjust your optimization workflow:

• Day 0-1: Make decisions based on leading indicators (click volume, CPM, CTR)
• Day 2-3: Validate with initial SKAN postbacks
• Day 7+: Make budget decisions based on accumulated SKAN data with enough volume for reliability

Don't optimize daily based on SKAN data. The randomized delays mean any single day's data is incomplete. Weekly optimization cycles align better with SKAN's data delivery patterns.

The bigger picture: iOS attribution in 2026

SKAN 5.0 is a step forward, but it doesn't return iOS measurement to pre-ATT levels. The reality is that iOS attribution will remain limited compared to Android and web for the foreseeable future.

Successful media buyers in 2026 operate with this framework:

1. SKAN for Apple-sanctioned install and re-engagement attribution
2. Server-side tracking for web-based conversion measurement
3. MMM or geo testing for overall channel incrementality validation
4. First-party data as the foundation for all measurement

No single data source tells the complete story. The brands and agencies that combine multiple measurement approaches -- and understand the strengths and limitations of each -- make the best budget decisions.

FAQ

Do I need to update my MMP integration for SKAN 5.0?

Yes. Your Mobile Measurement Partner (AppsFlyer, Adjust, Singular, Branch) needs to be updated to handle SKAN 5.0 postbacks, including re-engagement attribution and the fourth conversion window. Contact your MMP for their SKAN 5.0 migration guide.

How does SKAN 5.0 affect my Meta Ads iOS campaigns?

Meta's Ads Manager will incorporate SKAN 5.0 data into its reporting, providing better re-engagement measurement and longer conversion windows. However, Meta still relies heavily on its own modeled conversions for iOS campaigns, so SKAN data supplements rather than replaces Meta's reporting. The key improvement: you can now validate Meta's re-engagement claims against SKAN data.

Should I wait for SKAN 6.0?

No. Each SKAN version has been backwards-compatible, and the improvements are incremental. Implement SKAN 5.0 now and update when future versions arrive. Waiting for the "perfect" measurement framework means operating without any measurement in the meantime.

---

Go Funnel uses server-side tracking and multi-touch attribution to show you which ads actually drive revenue. Book a call to see your real numbers.