What Is SKAdNetwork? How iOS App Attribution Works ( Setup Guide)

...

Table of Contents

    MEET OUR TEAM!

    Summarize This Content with Artificial Intelligence (AI):

    If you run iOS user acquisition, you have almost certainly been frustrated by SKAdNetwork. Conversions arrive late, they arrive in aggregate rather than per user, some do not arrive at all, and the numbers rarely match what your other tools report. None of that is a bug. It is exactly how the framework is designed to behave, and once you understand why, iOS attribution stops looking broken and starts looking like a system with rules you can work within.

    SKAdNetwork, often shortened to SKAN, is Apple’s framework for measuring app install campaigns on iOS without tracking individual users. It is the privacy-preserving replacement for the user-level attribution that dominated mobile marketing before Apple’s privacy changes, and since those changes it has become the default way iOS install campaigns are measured. Understanding it is no longer optional for anyone spending on iOS.

    This guide explains what SKAdNetwork is, how its measurement flow actually works, what changed in the current version, how to set it up, and the mistakes that quietly corrupt the data. The goal is that by the end, the framework’s deliberate limitations read as constraints to plan around rather than failures to fix.

    What Is SKAdNetwork? A Clear Definition

    SKAdNetwork is Apple’s privacy-preserving attribution framework that measures which ad campaigns drive iOS app installs without exposing any individual user’s data. It works by putting Apple in the middle as a trusted intermediary. The ad network registers a campaign, the user installs the app, and Apple, rather than the advertiser or the ad network, validates the install and sends back a signed, aggregated, deliberately delayed report called a postback. No device identifier changes hands, and no user is tracked across apps.

    The framework exists because of Apple’s App Tracking Transparency policy, which requires apps to ask permission before tracking users across other companies’ apps and websites. When most users decline that prompt, the user-level tracking that mobile attribution once relied on disappears. SKAdNetwork is Apple’s answer to the question that created: how can advertisers still learn which campaigns work when they can no longer see what individual users do. The connection to the broader shift is covered in depth in the guide to app marketing attribution models, which places SKAN alongside the other measurement approaches a modern mobile program runs.

    The trade is explicit: privacy for the user in exchange for less granular data for the advertiser. SKAdNetwork gives you enough signal to tell which campaigns produce installs and, to a limited degree, what those users do afterward, but it does so in aggregate, with delay, and with rules designed to prevent any individual from being re-identified. That trade is the entire character of the framework, and every quirk that follows comes from it.

    Peaker Note: Why Your SKAN Numbers Never Match Your MMP

    The most common confusion Digipeak hears from iOS advertisers is that SKAdNetwork numbers do not reconcile with what their mobile measurement partner or in-app analytics report…

    How SKAdNetwork Works: The Measurement Flow

    The SKAdNetwork flow has four stages. Following them in order is the fastest way to understand why the data looks the way it does.

    Stage 1: Campaign registration. The ad network registers the campaign with Apple and serves the ad. When a user taps or views the ad, the ad carries a signed set of campaign parameters that Apple can later verify. This signature is what lets Apple trust that the install came from a real, registered campaign rather than a fabricated claim.

    Stage 2: Install and attribution. The user installs the app from the App Store. Apple, operating on the device and in its own systems, matches the install to the campaign that drove it. Crucially, this matching happens inside Apple’s trusted environment, not in the advertiser’s or the ad network’s systems, which is what keeps the process private.

    Stage 3: The conversion value. After the install, the app can record a small amount of post-install activity by updating a conversion value, a compact code that represents what the user did in their first hours or days (completed onboarding, made a purchase, reached a certain level). This is the one window the advertiser has into post-install behavior, and it is deliberately narrow so it cannot be used to fingerprint individuals.

    Stage 4: The postback. After a delay, Apple sends a signed postback to the ad network (and, if configured, a copy to the advertiser) reporting the install and the conversion value. The delay is intentional, and so is the fact that low-volume campaigns may receive their conversion value stripped out, because a conversion value attached to very few installs could theoretically identify someone. The postback is aggregated and timing-randomized precisely to prevent that.

    That flow, registration to install to conversion value to delayed aggregated postback, is the whole system. Every characteristic that frustrates advertisers (the delay, the aggregation, the missing conversion values on small campaigns) is a direct consequence of a stage in this flow doing its privacy job.

    What Changed in the Current Version of SKAdNetwork

    SKAdNetwork has evolved across versions, and the current generation added meaningful capability while keeping the privacy model intact. Three changes matter most for how campaigns are measured today.

    Multiple postbacks over time. Earlier SKAN sent a single postback, giving one narrow snapshot of post-install behavior. The current version sends multiple postbacks across successive time windows, which lets advertisers see not just the install but a coarse view of how engagement develops over the first days and weeks. This is a significant improvement for understanding whether a campaign drives users who stick, not just users who install.

    Coarse and fine conversion values. The current version introduced two grades of conversion value: a fine-grained value available when there is enough install volume to protect privacy, and a coarse value (low, medium, high) that can still be reported when volume is lower. This means smaller campaigns that previously received no conversion signal at all can now receive at least a coarse one, recovering measurement that was simply lost before.

    Crowd anonymity tiers. The amount of detail Apple returns now scales with campaign volume through a tiered system Apple calls crowd anonymity. Higher-volume campaigns clear the threshold for more detailed data because more installs make it harder to isolate any individual; lower-volume campaigns receive less detail. This is why consolidating spend into fewer, larger campaigns often produces better measurement than spreading it thin, a counterintuitive but important planning consequence.

    The through-line across all three changes: Apple is progressively giving advertisers more usable signal while never abandoning the principle that no individual can be re-identified. Each improvement is gated by volume thresholds that exist to preserve anonymity. Planning around those thresholds is now a core part of iOS measurement strategy.

    Struggling to get usable data out of SKAdNetwork on your iOS campaigns?

    Digipeak designs conversion value schemas, structures campaigns around crowd anonymity thresholds, and builds iOS measurement that actually informs optimization. Get in touch for a SKAN setup review.

    Get a SKAN Setup Review →

    Clicks but No Consultations?
    Turn Traffic into Appointments.

    Patient-first campaigns, A/B testing, and booking optimization to boost real consultations. Talk to our healthcare growth team.

    The Heart of SKAN: Conversion Value Mapping

    The conversion value is where most of the strategic work in SKAdNetwork lives, because it is the one part of the system the advertiser controls. Everything else is Apple’s; the conversion value schema is yours to design.

    A conversion value is a small code the app sets after install to represent what the user did. Because the space is limited, the schema (the mapping of codes to real behaviors) has to be designed deliberately. A schema that maps values to revenue tiers answers a different question than one that maps values to funnel milestones like onboarding completion or day-one retention. You cannot measure everything, so the schema forces a choice about what matters most to your business.

    This is the decision that separates a well-measured iOS program from a poorly-measured one. A team that maps conversion values thoughtfully to the events that predict long-term value gets a usable, if coarse, signal about campaign quality. A team that uses a default or arbitrary schema gets postbacks full of numbers that do not mean anything actionable. The framework will report faithfully whatever you tell it to report; designing what it should report is the work.

    AspectFine Conversion ValueCoarse Conversion Value
    Detail levelGranular (many possible values)Three tiers: low, medium, high
    When availableHigher-volume campaignsLower-volume campaigns
    Best forRevenue tiers, detailed funnelBroad quality signal
    Privacy basisRequires volume for anonymityUsable below the fine threshold
    Planning useOptimize on precise valueOptimize on directional quality

    How to Set Up SKAdNetwork: A Practical Walkthrough

    Setting up SKAdNetwork correctly runs through five stages. The order matters, because a mistake early (particularly in the conversion value schema) corrupts everything downstream.

    Step 1: Confirm the Framework Is Enabled in Your App

    SKAdNetwork support has to be present in the app build, with the relevant ad network identifiers included so campaigns from those networks can be attributed. Most apps running paid UA already have this through their mobile measurement partner’s SDK, but it is worth confirming rather than assuming, because an install from an unregistered network simply will not be attributed.

    Step 2: Design Your Conversion Value Schema

    This is the most important step and the one that rewards the most thought. Decide what post-install behaviors matter most (revenue, key funnel milestones, early retention) and map them to conversion values before any campaign runs. Design for the question you most need answered, because the schema determines what every postback can possibly tell you. Changing the schema later means losing comparability with earlier data, so getting it right at the start has outsized value.

    Step 3: Set the Measurement Window

    Decide how long after install the app should keep updating the conversion value. A short window captures immediate actions like onboarding; a longer window captures behaviors like a first purchase that may take days. The right window depends on how quickly your most valuable post-install behavior typically happens, and it is a genuine trade-off, because a longer window delays the postback further.

    Take Advantage of Automation with Artificial Intelligence!

    How can you use your time more efficiently? Artificial intelligence saves you time by automating repetitive tasks. Learn how you can leverage AI to accelerate your business processes.

      Step 4: Configure Multiple Postbacks (Current Version)

      Take advantage of the multiple-postback capability by configuring the successive time windows so you see how engagement develops rather than only the install moment. This is where the current version earns its value: a campaign that drives many installs but little day-seven engagement looks different from one that drives fewer, stickier users, and only multiple postbacks reveal the difference.

      Your Visuals Stand Out But You're Not Getting Any Feedback?
      Get a Free Design Audit Now!

      Let's evaluate together how you can better engage with your target audience visually, dysfunctional designs, and tips to increase engagement. Fill out the form now and get your free analysis report!

        Step 5: Plan Campaign Structure Around Crowd Anonymity

        Because the detail Apple returns scales with volume, structure campaigns so they clear the thresholds that unlock fuller data. In practice this usually means fewer, larger campaigns rather than many small ones, which runs against the old instinct to segment finely. Over-segmentation on iOS now actively costs you measurement detail, so campaign consolidation is a measurement decision as much as a media one.

        With those five in place, the framework is configured to give you the most usable signal it can within the privacy rules: attributed installs, a meaningful conversion value schema, an appropriate window, engagement development across multiple postbacks, and a campaign structure that clears the anonymity thresholds.

        Peaker Note: Consolidation Is a Measurement Strategy

        One counterintuitive lesson Digipeak applies on every iOS program: the old habit of slicing campaigns into many tightly-targeted segments now works against you…

        The Limitations You Have to Plan Around

        SKAdNetwork is workable, but only if you accept its constraints rather than fighting them. Four limitations shape every iOS measurement strategy.

        Delayed data. Postbacks arrive after a deliberate delay, so iOS campaign optimization cannot happen in real time the way it once did. Decisions are made on data that is hours or days old, which changes the optimization cadence from continuous to periodic. Teams used to same-day tuning have to adjust their rhythm.

        Aggregated, not user-level. SKAN tells you about campaigns, not people. You cannot follow an individual user’s journey through SKAN data, which means user-level analysis has to come from consented in-app analytics, a separate and smaller data set. The two coexist rather than combine.

        The crowd anonymity floor. Low-volume campaigns receive less detail or, below certain thresholds, lose the conversion value entirely. This structurally disadvantages small campaigns and niche audiences, and it is the single biggest reason to consolidate spend rather than fragment it on iOS.

        A limited behavioral window. The conversion value captures only a compressed slice of post-install behavior, so nuanced or long-horizon outcomes are hard to measure through SKAN alone. Understanding lifetime value on iOS requires modeling on top of the coarse signal rather than reading it directly.

        The realistic stance for 2026 is that SKAdNetwork is the privacy-safe backbone of iOS attribution, not a complete measurement solution on its own. Mature iOS programs pair it with consented analytics, incrementality testing, and modeling to build a fuller picture, treating SKAN as the trustworthy campaign-attribution layer rather than expecting it to answer every question.

        Frequently Asked Questions About SKAdNetwork

        What is SKAdNetwork in simple terms?

        SKAdNetwork is Apple’s framework for measuring which ad campaigns drive iOS app installs without tracking individual users. Apple sits in the middle as a trusted intermediary, validates the install, and sends back an aggregated, delayed report called a postback. It tells advertisers which campaigns work while keeping each user’s data private, which is the trade at the center of how it operates.

        Why did Apple create SKAdNetwork?

        Apple created SKAdNetwork to give advertisers a way to measure app install campaigns after its App Tracking Transparency policy required apps to ask permission before tracking users across other apps. When most users decline that prompt, traditional user-level attribution stops working. SKAdNetwork is the privacy-preserving replacement: it measures campaign performance in aggregate rather than following individuals.

        What is a conversion value in SKAdNetwork?

        A conversion value is a small code an app records after install to represent what the user did, such as completing onboarding, making a purchase, or reaching a milestone. It is the one window advertisers have into post-install behavior. Because the space is limited, designing the conversion value schema (which behaviors map to which codes) is the most important strategic decision in SKAdNetwork setup.

        Why does SKAdNetwork data not match my other analytics?

        Because they measure different things. SKAdNetwork reports aggregated, delayed, Apple-validated installs with a coarse conversion value, while in-app analytics report user-level behavior for measurable users. The two are built on opposite principles and will not reconcile. The productive approach is to use SKAN as the privacy-safe source for campaign attribution and analytics for richer behavioral detail, rather than trying to force one number from both.

        What is crowd anonymity in SKAdNetwork?

        Crowd anonymity is Apple’s system for scaling how much detail it returns based on campaign volume. Higher-volume campaigns clear thresholds for more detailed data because more installs make it harder to identify any individual, while lower-volume campaigns receive less detail or lose the conversion value entirely. This is why consolidating spend into fewer, larger campaigns often produces better measurement on iOS.

        Can I still optimize iOS campaigns in real time with SKAdNetwork?

        Not in the same way as before. SKAdNetwork postbacks arrive after a deliberate delay, so optimization happens on a periodic rather than continuous basis. Decisions are made on data that is hours or days old. Teams accustomed to same-day tuning need to adjust their cadence, and many pair SKAN with consented analytics and modeling to inform faster decisions where privacy allows.

        Is SKAdNetwork enough on its own for iOS measurement?

        No. SKAdNetwork is the privacy-safe backbone of iOS attribution, but mature programs pair it with consented in-app analytics, incrementality testing, and modeling to build a complete picture. SKAN answers which campaigns drive installs and gives a coarse post-install signal; the fuller questions about user behavior and lifetime value require additional measurement layered on top.

        Working With the Framework, Not Against It

        SKAdNetwork frustrates advertisers who expect it to behave like the user-level attribution it replaced, and rewards those who accept it on its own terms. The delayed, aggregated, privacy-first design is not a set of bugs to work around but the defining feature of the framework, and every quirk (the reconciliation gaps, the missing conversion values on small campaigns, the periodic rather than real-time optimization) follows directly from that design.

        Three takeaways for iOS teams: design the conversion value schema deliberately, because it is the one part of the system you control and it determines what every postback can tell you; consolidate campaigns to clear crowd anonymity thresholds, because on iOS fragmentation costs measurement detail; and treat SKAN as the attribution backbone rather than the whole solution, pairing it with consented analytics and modeling for the questions it cannot answer alone.

        The framework will keep evolving, and the direction is consistent: Apple gradually returns more usable signal while holding the line on individual privacy. The teams that build their iOS measurement around the framework’s logic now, rather than resenting its limits, are the ones positioned to benefit as each version widens what is measurable within the privacy model.

        How Does Digipeak Approach iOS Measurement and SKAdNetwork?

        At Digipeak, SKAdNetwork is treated as the foundation of a broader iOS measurement system rather than a box to check. Engagements begin by reviewing the conversion value schema, because a poorly designed schema is the most common reason an iOS program produces postbacks full of numbers that do not inform any decision. From there the work extends to campaign structure, ensuring spend is consolidated enough to clear crowd anonymity thresholds and unlock the detail the framework can provide, which connects directly to the ASO and app growth work that determines whether that measured traffic converts and retains.

        The measurement system Digipeak builds pairs SKAN with the other layers a mature iOS program needs, the same measurement discipline applied to server-side tracking on the web and to attribution modeling across channels. The principle is consistent across surfaces: privacy-safe measurement, designed deliberately, read honestly, and never expected to reconcile systems built on opposite foundations.

        Digipeak operates as a 360-degree growth agency from offices in London, Istanbul, and Texas, holds Google and Meta Partner status, and manages iOS measurement alongside the UA, ASO, and creative work that depends on it. If your SKAdNetwork data feels more like noise than signal, the conversion value schema is usually where the diagnosis starts.

        Share Content

        ...

        Join the
        Digipeak Newsletter!

        Stay up to date with the latest digital marketing insights, trends, and updates. Subscribe today!

          Related Articles