App Install Tracking Without IDFA in 2026

How does app install tracking work without IDFA in 2026?

Apple introduced App Tracking Transparency in 2021, effectively killing the most basic form of tracking we've ever known. Prior to ATT, every iPhone came with a persistent, unchanging code called an Identifier for Advertisers, which could be seen at both ends of the tracking process. You'd click a link, then the app would be installed. Now both devices know the same IDFA and the click is tied to that install. Very easy.

Not anymore. With ATT, users are prompted to agree to tracking. And they won't. Adjust's 2025 Global App Trends Report reveals that the average opt-in rate for all app categories was only 18%, down from an initial opt-in of 25% following the launch of ATT. With social and entertainment applications, that number drops to 12%. Not to mention you wouldn't blame them, either. "Allow this app to track your activity across other companies' apps and websites?" is a question nobody really wants to say yes to.

So now if you have an application that tracks installs from clicks on links, whether you're tracking campaign performance, assessing how your content creators are doing, or simply trying to know where your app users are coming from, you need a way to do this without the IDFA. Below is a snapshot of the current tracking landscape in 2026.

Why is install tracking a simpler problem than ad attribution?

When it comes to your own links, creator campaigns, or even email campaigns, install tracking and ad attribution are entirely distinct problems. In terms of ad attribution, there are likely a number of sources involved across multiple ad networks. In terms of install tracking from your own links, the process is much shorter:

• A tracking link is clicked.
• The app is downloaded and installed.
• The install is connected to the link clicked.

This means you don't have to take into account multiple ad views, view-through windows, or attribution models. Usually it will simply be a click then an install. According to AppsFlyer's 2025 attribution report, the median number of touchpoints in ad attribution campaigns is 4.2. But it'll almost always just be one touch point with tracking links. Because of this, we can achieve similar high tracking accuracy without as complicated of a system. We don't necessarily need a full-blown MMP.

How does Android handle install tracking?

Android has the best solution in the industry for passing referral data through the install process. The Install Referrer API passes this information along from the moment the link is clicked, through the Play Store install process, to the app, so the app has access to the information as soon as it is launched. You know exactly where the install came from without guessing.

This technology has been reliable and stable since its release. According to Google's 2025 Developer documentation, referrer data will remain valid for a period of 90 days from the click. Install referrer is supported on practically all Android devices that run Play Services, which according to Statista makes up 95% of the currently installed base of Android.

It's not a flawless system, however. If the user clicks a link and then goes to manually search for and install an app, the referral information will not pass. If the user has installed apps on another device, the referral won't pass. For the majority of users who install your app by clicking on a link, the process works exactly as intended. Most tracking providers have an attribution accuracy for Android click-to-install between 85-90% in practice, which is enough to launch a proper application.

How does iOS tracking work without IDFA?

iOS presents a tougher challenge. We don’t have a counterpart to Android’s Install Referrer API and the user’s link-clicked browser session is totally unconnected to the app install session. Your solution must tie these together in privacy-respecting ways that don’t use IDFA.

Today, iOS install tracking connects clicks to installs by matching privacy-preserving device signals that are available without requiring special permissions or asking users to accept tracking. No need for the ATT prompt, IDFA, or app-to-app tracking. The result is precise enough to reliably track installs from your own links.

Apple has spent the past several years closing each tracking hole they could identify. ITP from Safari has blocked 3rd party cookies from functioning in Safari, iCloud+ users (that Apple now says are over 130M as of early 2025) benefit from Private Relay, and with each new release Apple continues to shut more signals down. The 2026 iOS tracking winners will be those that worked within the limits instead of fighting around them.

Why doesn’t SKAdNetwork work for link tracking?

SKAdNetwork was designed for ad networks, not for you tracking your own links. The issues:

• Not granular to a single link: SKAN will show that it got 50 installs from TikTok but can’t show them to your different tracking links
• Limited to 100 campaign IDs: Won’t be enough for creators or campaigns
• 24-48hr minimum delay: Postback is delayed, and Apple adds a randomized delay
• Single Conversion Value (0-63): Not enough resolution to measure revenue
• No real-time data: Creators and marketers need to see results in real time

SKAN is built for ad network attribution at the network level and doesn’t support tracking of individual links or creators. Skip it.

What to look for in a post-IDFA tracking solution

When you’re choosing an install attribution tool in 2026, focus on:

• Doesn’t require IDFA. A tracking solution that needs IDFA to function will cost you App Store installs and conversion. You’re getting people who have the IDFA permission and are asking them to download your app. A 2025 Storemaven report found that showing the ATT during an app’s onboarding process reduced Day 1 retention by between 8-15% based on app category. A good install tracking tool won’t need it.

• Works on both iOS and Android. Your app is probably available on both (it probably should be). Choose a tracking tool that can track both with a single SDK and consolidated reporting. 78% of the top 1,000 highest revenue apps are on both, per Statista (2025).

• Per-link tracking. It's not just "which channel" or "which campaign" that led to which install, it's which specific link. This granularity is what you need to know to truly measure performance and optimize budget.

• Real-time reporting. If you're working with creators, they want to see the results as they happen. Slow reporting is momentum-killing. A 2025 CreatorIQ study found that creators who got same-day feedback were 2.3x as likely to produce more content that same week.

• Revenue tracking. Install tracking tells you who drives installs. Revenue tracking tells you who drives paying users. If you care about ROI on your marketing investment, you need to connect installs to actual purchases. Connect RevenueCat, Stripe, Superwall, or Adapty and see revenue per link.

• Lightweight SDK. The tracking SDK should add as little weight as possible to your app, requiring the minimal number of permissions. You need an SDK under 50KB with a minimum of dependencies. An SDK that adds 5MB to your app will increase startup time and make privacy-sensitive users nervous.

• Honest accuracy claims. If a platform is claiming 100% accurate attribution on iOS without IDFA, then they're either lying to you or doing things that will break eventually. Look for transparency about what can't be tracked and what can.

The hard truths of post-IDFA tracking

No-IDFA tracking is a constraint, not a feature. Yes, not requiring the ATT prompt is nicer onboarding and not tracking device IDs is a simpler privacy story, but nobody should say that this is better than matching with full user consent. We're just doing the best we can in the constraints of reality.

The Android approach is good. The Android referral flow is pretty stable and well documented. iOS is a little more shaky. The platforms that fell off after iOS 14.5 were the ones that didn't start building for a privacy-first approach then. The same will happen with future iOS releases.

There is some traffic that you just can't track. Users on VPNs, users that wait a long time after the click to actually install, users behind shared networks with many similar devices. The question is never whether you'll have gaps in your tracking, but whether the percentage of your traffic that is trackable will give you enough actionable intelligence to actually understand where your installs are coming from: which links are working, which creators are performing, what your real installs are.

And in practice it is. But it's also not perfect. What's quietly interesting to me about the post-IDFA tracking world is that it's the ones that have been tracking in a lazy way that have gotten punished the most. The former approach was: grab device ID and match across context. It required very little thought. The tracking folks that do tracking really well now are the ones that actually sat down and thought about what signals matter for what their specific use case is versus trying to grab a universal identifier that never belonged to them in the first place.