A reattribution window is a set period of time after someone installs an app during which a reinstall (or reopen via an ad) can still be counted as a “new” attributed conversion. Mobile measurement partners (MMPs) focused on install-only reporting typically use reattribution windows to give credit to campaigns driving reengagement via app redownloads.

Where they do apply

Reattribution windows only apply to install-touch attribution — the act of reassigning credit to an ad for driving a new install after a previous uninstall. Marketers designed them for a world where install volume was the primary goal, not long-term user growth.

Where they do not apply

Reattribution windows do not apply to last-touch attribution, nor do they apply to non-install conversion events like purchases, subscriptions, or sign-ups. In Branch’s model — where we continuously measure user activity across the full lifecycle — there is no arbitrary cutoff where users “reset” or need to be treated as new.

The problem with reattribution windows in modern growth strategies

Modern marketers optimize beyond installs. Tying attribution logic to reinstall events within a predefined time window fragments insight, obscures user intent, and undervalues post-install influence. These windows often ignore critical activity across channels and devices, cutting marketers off from the full user journey.

Branch customers look for more value out of their broader marketing efforts.

Branch’s stance

Branch operates under the principle of lifecycle-based attribution, not windowed event recapture. We measure influence from the first touch to the final conversion — across installs, reengagements, and revenue-driving events — without artificially resetting or gating attribution behind arbitrary reattribution windows.

• Last-touch attribution remains valid across a user’s life, not just during an install window.
• Every conversion — install or not — has a real path and real marketing influence behind it.
• You don’t need a reattribution window when your attribution is already continuous and durable.

Who this matters for

If you only care about app installs, reattribution windows might give you comfort — but not insight. If you’re focused on full-funnel growth, retention, and monetization, you need attribution that reflects the entire user journey, not one that resets arbitrarily.

Summary

Reattribution windows are a relic of install-only marketing. They don’t apply to last-touch attribution and are incompatible with today’s sustainable growth strategies focused on engagement and value. Branch’s attribution model provides durable identity, cross-channel context, and continuous measurement — without the need for outdated windows.

Connect with us if you have any questions. We’re always here to help!

The post Why Reattribution Windows Are Misunderstood — and Not a Fit for Last-Touch or Full-Funnel Measurement appeared first on Branch.

Let’s dig in.

The known: AdAttributionKit grows up

AdAttributionKit (AAK), Apple’s SKAdNetwork successor, got a long-awaited set of upgrades that push it closer to being a partial multi-touch attribution (MTA) framework. No, it’s not a full MTA — but it’s starting to look like one in spirit.

Here’s what’s new and why it matters:

• Conversion Tags
  Apple now lets advertisers append a conversion-tag to deep links. You pull this into your app and tie it to updateConversionValue, giving you more control over who gets credit. It’s a clever way of letting the ecosystem influence credit without breaking SKAN’s rules.
• Custom Attribution Rules
  Finally — configurable attribution windows! Developers can now set these windows per network or globally, directly in the app’s Info.plist. The default 30d click / 1d view model is no longer forced. You can even exclude view-through altogether.
• Attribution Cooldown (Reattribution, Reimagined)
  This one’s nuanced. Instead of a reattribution window, Apple now lets devs block overlapping post-install attributions for a set number of hours after a conversion. It’s like saying: “Hold off on awarding new credit just yet.” This is useful for preventing hijacked credit on rapid-fire conversions from reengagement strategies.
• Geo in Postbacks
  Postbacks now include country code — sourced from device settings. But only for countries with alternative app stores (think EU DMA regions), so don’t expect this to unlock geo-targeting at scale (yet). Still, it’s a notable expansion of the data allowed.

While Apple isn’t offering true MTA, features like Conversion Tags and Attribution Cooldown signal a meaningful shift in that direction. Conversion Tags allow advertisers to decide who gets credit by passing internal context into the updateConversionValue() call — essentially letting brands override last-touch based on known prior engagement. Attribution Cooldown, meanwhile, protects earlier touchpoints by blocking late signals from stealing credit for a defined time window. These tools don’t expose multiple touchpoints or enable fractional credit, but they empower advertisers to apply MTA-like logic within Apple’s sandbox.

The understated: fingerprinting protections get serious

Apple quietly teased a major privacy enhancement in iOS 26:

“Browsing in Safari gets even more private with advanced fingerprinting protection extending to all browsing by default.”

At first glance, this sounds like more of Apple’s usual privacy rhetoric. But under the hood, it’s a continuation — and escalation — of a broader strategy already in motion. Safari 26 builds directly on Advanced Tracking and Fingerprinting Protection (ATFP) and Link Tracking Protection (LTP), pushing both deeper into the browser stack.

According to the latest WebKit update, Safari will now:

• Block known fingerprinting scripts from accessing sensitive APIs (e.g., screen dimensions, canvas, Apple Pay status, web audio)
• Prevent these scripts from using long-lived client-side storage like cookies and localStorage
• Strip access to navigational state like document.referrer and query parameters — the foundation of many click-based measurement flows

Apple isn’t using a catch-all approach — they’re targeting “known” trackers via curated lists. But which list? Apple already maintains a Privacy Impacting SDKs list, as well as one for known tracking query parameters used in LTP. It’s unclear if this is a new list, a consolidation, or an expansion — but it’s clear that Safari is now enforcing privacy on-device with more precision and fewer workarounds.

So what exactly is affected?

Apple appears to be using curated blocklists of “known trackers,” drawing from sources like PrivacyTests.org — which includes identifiers like gclid, fbclid, and others common in paid media. This is an extension of Link Tracking Protection (LTP) and Advanced Tracking & Fingerprinting Protection (ATFP), targeting invasive third-party click-based tracking.

What this could mean

As usual Apple isn’t 100% clear on how these policies will take shape, but given the information available, this is how we think things will shake out:

• Deep linking and owned attribution are safe
  Branch deep linking and deferred deep linking are not impacted. Our query parameters are not on Apple’s known-blocked list. That means owned, organic, and non-paid traffic measurement remains fully intact.
• Paid media click IDs are vulnerable
  If your attribution model depends on click identifiers like gclid, fbclid, or other query-level tracking parameters, those may be scrubbed by ATFP in Safari — especially for paid campaigns. This will likely affect deterministic link-based attribution for browser-based flows.
• Modeled attribution is more relevant than ever
  While this update sounds dramatic, its real-world impact is softened by the industry’s shift toward probabilistic and modeled attribution. At Branch, we’ve already been investing here. As click IDs become less reliable, modeled attribution will increasingly carry the weight.

Apple’s privacy agenda is marching forward — but not everything is breaking. What we’re seeing is a shift in where deterministic measurement applies, and an acceleration toward smarter, privacy-preserving alternatives.

TL;DR

Apple continues its long game of reshaping attribution — tightening browser-level privacy while offering more flexibility inside its own ecosystem. AdAttributionKit gets more powerful, letting developers define attribution windows, control re-engagement logic, and even influence credit assignment via Conversion Tags.

Meanwhile, Safari’s fingerprinting protections expand — but enforcement remains in early stages.

For Branch and our customers, this reaffirms a few key bets:

• Modeled attribution (like Predictive Aggregate Measurement) — probabilistic, and privacy-safe — is not a fallback, it’s the future.
• Deterministic identity in the browser is getting harder, fast.
• The browser is no longer neutral infrastructure — it’s an opinionated gatekeeper.

Let’s connect if you want to unpack how these shifts might impact your measurement stack, your partners, or your roadmap.

The post WWDC 2025: The Shiny, The Sharp, and the Subtle Shifts in Attribution and Privacy appeared first on Branch.

Let’s dig in.

The known: AdAttributionKit grows up

AdAttributionKit (AAK), Apple’s SKAdNetwork successor, got a long-awaited set of upgrades that push it closer to being a partial multi-touch attribution (MTA) framework. No, it’s not a full MTA — but it’s starting to look like one in spirit.

Here’s what’s new and why it matters:

• Conversion Tags
  Apple now lets advertisers append a conversion-tag to deep links. You pull this into your app and tie it to updateConversionValue, giving you more control over who gets credit. It’s a clever way of letting the ecosystem influence credit without breaking SKAN’s rules.
• Custom Attribution Rules
  Finally — configurable attribution windows! Developers can now set these windows per network or globally, directly in the app’s Info.plist. The default 30d click / 1d view model is no longer forced. You can even exclude view-through altogether.
• Attribution Cooldown (Reattribution, Reimagined)
  This one’s nuanced. Instead of a reattribution window, Apple now lets devs block overlapping post-install attributions for a set number of hours after a conversion. It’s like saying: “Hold off on awarding new credit just yet.” This is useful for preventing hijacked credit on rapid-fire conversions from reengagement strategies.
• Geo in Postbacks
  Postbacks now include country code — sourced from device settings. But only for countries with alternative app stores (think EU DMA regions), so don’t expect this to unlock geo-targeting at scale (yet). Still, it’s a notable expansion of the data allowed.

While Apple isn’t offering true MTA, features like Conversion Tags and Attribution Cooldown signal a meaningful shift in that direction. Conversion Tags allow advertisers to decide who gets credit by passing internal context into the updateConversionValue() call — essentially letting brands override last-touch based on known prior engagement. Attribution Cooldown, meanwhile, protects earlier touchpoints by blocking late signals from stealing credit for a defined time window. These tools don’t expose multiple touchpoints or enable fractional credit, but they empower advertisers to apply MTA-like logic within Apple’s sandbox.

The understated: fingerprinting protections get serious

Apple quietly teased a major privacy enhancement in iOS 26:

“Browsing in Safari gets even more private with advanced fingerprinting protection extending to all browsing by default.”

At first glance, this sounds like more of Apple’s usual privacy rhetoric. But under the hood, it’s a continuation — and escalation — of a broader strategy already in motion. Safari 26 builds directly on Advanced Tracking and Fingerprinting Protection (ATFP) and Link Tracking Protection (LTP), pushing both deeper into the browser stack.

According to the latest WebKit update, Safari will now:

• Block known fingerprinting scripts from accessing sensitive APIs (e.g., screen dimensions, canvas, Apple Pay status, web audio)
• Prevent these scripts from using long-lived client-side storage like cookies and localStorage
• Strip access to navigational state like document.referrer and query parameters — the foundation of many click-based measurement flows

Apple isn’t using a catch-all approach — they’re targeting “known” trackers via curated lists. But which list? Apple already maintains a Privacy Impacting SDKs list, as well as one for known tracking query parameters used in LTP. It’s unclear if this is a new list, a consolidation, or an expansion — but it’s clear that Safari is now enforcing privacy on-device with more precision and fewer workarounds.

So what exactly is affected?

Apple appears to be using curated blocklists of “known trackers,” drawing from sources like PrivacyTests.org — which includes identifiers like gclid, fbclid, and others common in paid media. This is an extension of Link Tracking Protection (LTP) and Advanced Tracking & Fingerprinting Protection (ATFP), targeting invasive third-party click-based tracking.

What this could mean

As usual Apple isn’t 100% clear on how these policies will take shape, but given the information available, this is how we think things will shake out:

• Deep linking and owned attribution are safe
  Branch deep linking and deferred deep linking are not impacted. Our query parameters are not on Apple’s known-blocked list. That means owned, organic, and non-paid traffic measurement remains fully intact.
• Paid media click IDs are vulnerable
  If your attribution model depends on click identifiers like gclid, fbclid, or other query-level tracking parameters, those may be scrubbed by ATFP in Safari — especially for paid campaigns. This will likely affect deterministic link-based attribution for browser-based flows.
• Modeled attribution is more relevant than ever
  While this update sounds dramatic, its real-world impact is softened by the industry’s shift toward probabilistic and modeled attribution. At Branch, we’ve already been investing here. As click IDs become less reliable, modeled attribution will increasingly carry the weight.

Apple’s privacy agenda is marching forward — but not everything is breaking. What we’re seeing is a shift in where deterministic measurement applies, and an acceleration toward smarter, privacy-preserving alternatives.

TL;DR

Apple continues its long game of reshaping attribution — tightening browser-level privacy while offering more flexibility inside its own ecosystem. AdAttributionKit gets more powerful, letting developers define attribution windows, control re-engagement logic, and even influence credit assignment via Conversion Tags.

Meanwhile, Safari’s fingerprinting protections expand — but enforcement remains in early stages.

For Branch and our customers, this reaffirms a few key bets:

• Modeled attribution (like Predictive Aggregate Measurement) — probabilistic, and privacy-safe — is not a fallback, it’s the future.
• Deterministic identity in the browser is getting harder, fast.
• The browser is no longer neutral infrastructure — it’s an opinionated gatekeeper.

Let’s connect if you want to unpack how these shifts might impact your measurement stack, your partners, or your roadmap.

The post WWDC 2025: <br>The Shiny, The Sharp, and the Subtle Shifts in Attribution and Privacy appeared first on Branch.

If you’re responsible for marketing strategies and trying to make sense of this rapidly evolving space, you’re not alone. Let’s break down what ICM is, why it matters, and how it fits into the broader narrative that includes solutions like Meta’s Aggregate Event Measurement (AEM) and Branch’s Predictive Aggregate Measurement (PAM).

What is Integrated Conversion Measurement?

Integrated Conversion Measurement (ICM) is Google’s latest solution aimed at enhanced conversion tracking and optimization for App campaigns across both Android and iOS platforms. As platform and regulatory changes have made it more challenging to understand the full impact and performance of App campaigns, ICM addresses these challenges by offering more accurate and comprehensive cross-channel attribution in MMP reporting.

ICM provides several key benefits:

• Enhanced reporting across iOS & Android: ICM enables real-time and event data reporting of campaign performance. On iOS, it offers improved reporting for mobile app conversions from iOS 14+ users. On Android, it enhances reporting for mobile app conversions from users in the European Economic Area (EEA) or when a user opts out of device-level permissions.
• Privacy-preserving techniques: Incorporates innovative technologies, such as on-device conversion measurement (ODM), to improve measurement accuracy, all without compromising user privacy.
• Flexible integrations: ICM integrates reporting improvements for advertisers using third-party App Attribution Partners, like Branch, and for those utilizing the App Conversion Tracking API for attribution.

Why ICM matters

Historically, Google Ads-driven, opted-out installs have been a visibility black hole. Advertisers have long struggled to understand how clicks on app-based ad inventory drive conversion actions on mobile especially under newer privacy regulations and platform policies (ATT).

ICM represents a new default future — one where deterministic, user-data level  is no longer feasible at scale. And Google is signaling that measurement must evolve or die.

This isn’t just a technical update. It’s a continuation of the fundamental rethinking of how performance marketing is measured:

• From people-based to aggregated attribution
• From deterministic to probabilistic and predictive models
• From granular user journeys to privacy-preserving conversion modeling

How it works compared to Meta AEM and Branch PAM

ICM isn’t launching into a vacuum. Let’s look at how it stacks up against other major attribution shifts:

While each solves different facets of the attribution problem, they’re all part of the same larger trend: modeled measurement is becoming the norm.

Branch’s PAM plays a critical role here by extending attribution beyond the SANs. Where ICM or AEM might capture one part of the journey, PAM helps stitch the broader story together across all marketing channels.

Mobile attribution has changed — Here’s the timeline

Mobile attribution milestone timeline

Below is a quick view of just how much has changed in the last ~20 months:

How to start planning for ICM now

1. Invest in cross-channel attribution tools: ICM will help, but only within the Google ecosystem. You need complementary solutions (like PAM) to analyze influence across TikTok, Snap, CTV, and more.
2. Branch will automatically support and begin reporting on ICM upon Google’s rollout.
   1. To get the latest performance benefits from on-device conversion measurement, update to the latest iOS SDK v3.11.0 to support integration with ODM.
3. Rethink KPIs: You’ll increasingly rely on aggregated and modeled data. Build trust in your modeled results now, so you’re not left scrambling.

Final thoughts

Google’s ICM isn’t just another adtech acronym. It’s a signpost for where mobile attribution and metrics are going: privacy-forward, aggregated, and increasingly predictive. Paired with tools like Meta AEM and Branch PAM, it paints a picture of a new attribution era that demands both technical flexibility and strategic foresight.

The marketers who thrive in this new era won’t be those chasing perfect data. They’ll be the ones adapting fastest to imperfect but actionable insights.

Have questions about mobile attribution? We’re here to help.

The post Google’s Integrated Conversion Measurement Confirms It: The Attribution Game Has Changed appeared first on Branch.