Support Home > Integrations > SKAdNetwork Integration

SKAdNetwork Integration

Feature Summary: Apple, via their SKAdNetwork framework, will now ingest ad signals and conversion signals, and report back anonymized deterministic attribution data. The following document describes the setup requirements for advertisers to prepare their apps, configure their conversion model, and start viewing SKAdNetwork data in the Kochava dashboard.


SKAdNetwork Certified Partners

To see a list of SKADNetwork certified partners, please refer to the following support document. If you are running or plan to run campaigns through SKAdNetwork with a partner that is not listed, please refer them to the following integration support document.


Prepare Your App to Report Conversions to SKAdNetwork

If Leveraging the Kochava SDK:

Implement minimum SDK version:

 

mParticle/Kochava Hybrid SDK Kit:

If you are currently leveraging the mParticle/Kochava hybrid SDK kit, and only plan to track install conversions (registerAppForAdNetworkAttribution) in the SKAdNetwork framework, implement the hybrid SDK kit with minimum Kochava iOS 4.1.1. SDK version.

If you plan to leverage the mParticle/Kochava hybrid SDK kit for conversion model support (includes updateConversionValue), please follow the steps outlined in this support doc. Implement the hybrid SDK kit with minimum Kochava iOS 4.2.0. SDK version. With the hybrid kit solution, installs are syndicated via the Kochava SDK, while events are syndicated via the mParticle SDK. For SKAdNetwork tracking, you will need to configure identical events in the Kochava SDK and use a boolean to indicate that the Kochava SDK events should not be sent off the device. This will ensure that the Kochava SDK can utilize the events for conversion model support, while preventing duplication of event feeds.


Configure Your SKAdNetwork Settings in the Kochava Dashboard

  1. Log in to Kochava.
  2. Select the desired Account and App.
  3. Select Integrations > SKAdNetwork.
  4.  


     

  5. Check Enable SKAdNetwork Reporting.
  6. Select the Method of Sending Conversion Data to SKAdNetwork:
    • Kochava SDK
    • NOTE: mParticle/Kochava hybrid SDK kit users should select the Kochava SDK.

    • Facebook SDK (Beta)
    • Other
    • NOTE: If you are not leveraging the Kochava SDK, select Other as your Method of Sending Conversion Data to SKAdNetwork. This will allow you to manually enter the meaning of each conversion value. Kochava will use these mappings to populate relevant fields in reporting.

     

    BETA: Kochava has built an integration with Facebook for clients who are leveraging the Facebook SDK for conversion model support. When selecting Facebook SDK as your Method of Sending Conversion Data to SKAdNetwork, Kochava will make a request to Facebook and automatically pull in your conversion value mappings, which will in turn be used to populate relevant fields in reporting. If changes are made in the Facebook dashboard to your conversion model, Kochava will automatically update your mappings. There is no need to return to this page after the initial setup.

     

    WARNING: Only one party should send conversion data to SKAdNetwork for a given app. If multiple parties are reporting conversions to SKAdNetwork, the signals will interfere with one another, leading to inaccurate conversion value data.

     

  7. Enter the App Store App ID.
  8. NOTE: The same app store app ID can only be entered on one Kochava app. If you are currently tracking an app with the same app store ID under multiple Kochava apps, please contact your Client Success Manager for further information.

  9. Select the Maximum Measurement Window:
    • Install Only
    • Day of Install
    • Day of Install + 1 Day
    • Day of Install + 2 Days
    • Day of Install + 3 Days
    • Day of Install + 4 Days
    • Day of Install + 5 Days
    • Day of Install + 6 Days
    • Day of Install + 7 Days

    NOTE: SKAdNetwork provides one postback per user (except redownloads). The measurement window determines the maximum duration to track a user’s activity before allowing SKAdNetwork to generate a postback. A shorter measurement window will allow for tracking more granular insights about the user’s activity.

  10. Select the Extension Interval:
  11. NOTE: Designate the window of opportunity for the SDK to trigger daily silent conversion updates when the app is open, to help extend Apple’s measurement window. Allowing more hours per day increases the chance users can be monitored closer to your maximum measurement window, but as a trade-off can also broaden the definition of the user’s measurement window (D0, D1, etc.) in reporting. This functionality is especially impactful in keeping the Apple window open in cases where users are sparsely triggering your conversion model target actions.

    • 12-24
    • 23-24

    Apple closes the SKAdNetwork measurement window 24 hours from the last SKAdNetwork API call to either updateConversionValue(_:) or registerAppForAdNetworkAttribution(). Once Apple closes the SKAdNetwork measurement window, future SKAdNetwork API calls are ignored by Apple and further SKAdNetwork measurement will not take place.

    If you are only interested in SKAdNetwork measurement within the first 24 hours following the install, these limitations aren’t a problem. However, if you’d like for SKAdNetwork measurement to extend beyond the first 24 hours and even up to Kochava’s maximum 7 day measurement period, these limitations make it difficult to measure the entire duration because if the user does not open the app or perform a conversion update during any 24 hour period during those 7 days, Apple will close the SKAdNetwork measurement window because 24 hours will have elapsed from the last SKAdNetwork API call.

    In order to keep the SKAdNetwork measurement window open, conversion updates must be triggered at regular intervals throughout the desired measurement period. Your conversion events triggered by the user will help to accomplish this, but these events alone may not always trigger at least once every 24 hours during the entire desired measurement period, which means the SKAdNetwork measurement window may close early.

    The Kochava SDK helps extend the SKAdNetwork measurement window during periods of user inactivity by triggering silent conversion updates at regular intervals. The frequency of these silent conversion updates is referred to as the extension interval, and may be either 23-24 hours or 12-24 hours. You can choose which interval to use under the Extension Interval option within your SKAdNetwork conversion model.

    Below is an explanation of these two options:

     

    Option A: 23-24 Hour Extension Interval:

    The SDK will trigger a silent conversion update as early as 23 hours, but no later than 24 hours since the last silent conversion update, and only if the app is open during that time. This option results in a more granular representation of the D value in terms of reporting, as each D value represents roughly a 24 hour period. However, the SDK is restricted to only 1 hour per day at which the app must be open to trigger a silent conversion update. This results in less opportunity for the SDK to help keep the SKAdNetwork measurement window open.

     

    Advantages —

    Reporting granularity of 24 hours per D value.

     

    Disadvantages —

    Limited opportunity for the SDK to help keep the SKAdNetwork measurement window open.

     

    Use Case —

    Useful when frequent conversion events are likely to be triggered at least once every 24 hours, as those events will often keep the SKAdNetwork measurement window open on their own without the help of the SDK.

     

    D Value Representation —

    Using the 23-24 hour option, each D value represents roughly 1 day.

    D Value Hour Since Install Day Since Install
    D0 0-24 hours 0
    D1 23-48 hours 1
    D2 46-72 hours 2
    D3 69-96 hours 3
    D4 92-120 hours 4
    D5 115-144 hours 5
    D6 148-168 hours 6
    D7 159-192 hours 7

    Option B: 12-24 Hour Extension Interval:

    The SDK will trigger a silent conversion update as early as 12 hours, but no later than 24 hours since the last silent conversion update, and only if the app is open during that time. This option results in a broader definition of the D value in terms of reporting, as each D value represents anywhere from a 12 to 24 hour period. As a trade off, the SDK is allowed 12 hours per day at which the app may be open to trigger a silent conversion update, which results in significantly more opportunities for the SDK to help keep the SKAdNetwork measurement window open.

     

    Advantages —

    Greater opportunity for the SDK to help keep the SKAdNetwork measurement window open for the entire desired measurement window.

     

    Disadvantages —

    Reporting granularity of 12-24 hours per D value, which broadens with each passing day.

     

    Use Case —

    Useful when conversion events are not likely to be regularly triggered at least once every 24 hours, as those conversion events will not likely be enough to keep the SKAdNetwork measurement window open on their own.

     

    D Value Representation —

    Using the 12-24 hour option, each D value represents as much as 1 day, but can also represent as little as half that time. This variance is based solely on how often the user opens the app. For users who open the app very frequently, the D value will tend to represent the lower limit of the range. For users who open the app infrequently, the D value will tend to represent the upper limit of the range.

    D Value Hour Since Install Day Since Install
    D0 0-24 hours 0
    D1 12-48 hours 0-1
    D2 424-72 hours 1-2
    D3 36-96 hours 1-3
    D4 48-120 hours 2-4
    D5 60-144 hours 2-5
    D6 72-168 hours 3-6
    D7 84-192 hours 3-7

    FAQ:


     

  12. Select the Conversion Model:
    • Highest Value Event
    • Revenue
    • User Journey
    • Engagement

     


    • User performance can be viewed based on the highest-value event completed by the user within the measurement window.

      The number of events that can be selected and tracked is dynamically adjusted based on the maximum measurement window selected.

       

      Max Measurement Window Maximum Events
      D0, D1 31 + 1 reserved for no events
      D2, D3 15 + 1 reserved for no events
      D4, D5, D6, D7 7 + 1 reserved for no events

       

      1. Click + to open the event menu.
      2. Select the desired event(s).
      3. Drag and drop events into the desired order.
      4. NOTE: Event 1 being the lowest value event.

       

      A. Click + to add Event(s).
      B. Events may be reordered by dragging and dropping an event into a new position.
      C. Click X to remove Event(s).


    • User performance can be viewed based on how much revenue the user generated within the measurement window. Summation of revenue will be based on values passed under the price event metadata parameters.

      The number of revenue buckets that can be set and tracked are dynamically adjusted based on the maximum measurement window selected.

       

      Max Measurement Window Maximum Revenue Buckets
      D0, D1 31 + 1 reserved for $0
      D2, D3 15 + 1 reserved for $0
      D4, D5, D6, D7 7 + 1 reserved for $0

       

      Revenue Buckets:

      NOTE: A minimum of 5 Revenue Buckets are required.

      1. Select Custom Ranges or Equal Increments.
      2.  

        Custom Ranges —

        NOTE: Bucket 1 will always be set at $0.00.

        1. Enter a Less Than value for Bucket 2.
        2. NOTE: Bucket 3 Greater Than value will be automatically filled by the value entered for Bucket 2.

        3. Click Add a Bucket.
        4. Enter the Less Than values for any additional buckets.

         

        Equal Increments —

        1. Enter an Increment.
        2. NOTE: Buckets will be automatically setup with the increment value entered.

       

      Revenue Event Filter:

      By default the model will aggregate revenue for any events marked as Revenue on the Apps & Assets > Event Manager page. Users can filter to specific revenue events if they wish to.

      1. Select All Events or Specific Events.
      2. If Specific Events has been selected:
        1. Click + to add an event.
        2. Click X to remove an event.

       

      Event Metadata Filter:

      When using Specific Key and Specific Key and Values, an event will only be included if it contains the selected metadata key, or the selected metadata key+value(s), respectively. Revenue keys, such as price and sum do not need to be added as filters.

      1. Select All Keys and Values, Specific Key or Specific Key and Values.
      2. Select the Key. (optional)
      3. Enter the Value. (optional>

       

      A. Enter Custom Ranges for Revenue Buckets.
      B. Enter Equal Increments for Revenue Buckets.
      C. Select either All Revenue Events or select specific Revenue Events.
      D. Select all or specific keys and values.


    • User performance can be viewed based on a list of events completed by the user, regardless of order, within the measurement window.

      The number of events that can be selected and tracked are dynamically adjusted based on the maximum measurement window selected.

       

      Max Measurement Window Maximum Revenue Events
      D0, D1 5
      D2, D3 4
      D4, D5, D6, D7 3

       

      1. Click + to add up to Three (3) events.
      2. Click X to remove an event.

       

      A. Click + to add a event(s).
      B. Click X to remove an event.


    • User performance can be viewed based on the number of times the user completed a specific event within the measurement window.

      The maximum number of times the event completion can be counted is dynamically adjusted based on the maximum measurement window selected.

       

      Max Measurement Window Maximum Number of Times Counted
      D0, D1 31 + 1 reserved for 0 times
      D2, D3 15 + 1 reserved for 0 times
      D4, D5, D6, D7 7 + 1 reserved for 0 times

       

      1. Select the desired Engagement Event.

       

     

  13. Click Save.

Sharing your Conversion Model Mappings (Conversion Model API)

Once you have configured your SKAdNetwork Settings as described above, your conversion model mappings can be made available to your media partners. Your partners will be the direct recipients of attribution postbacks from SKAdNetwork. These postbacks contain a conversion value 0-63 to indicate a users performance, but lack the associated meaning, which may be a revenue range, a specific event/list of events the user completed, or the number of times the user completed a target event. Kochava has made a Conversion Model API available for partners to access your mappings, which will allow them to interpret the meaning of the conversion value and utilize it for optimization or other purposes.

 

There are two prerequisites for a media partner to access your app’s mappings. They must be integrated with Kochava for SKAdNetwork support (see SKAdNetwork Certification), and the partner must be added in Partner Configuration in the Kochava dashboard, within the relevant app. If you are already working with this partner and have configured postbacks for them through Partner Configuration, the latter step is already completed.

NOTE: If you are using a Kochava conversion model, you can indicate that in the Facebook dashboard. Facebook will ask you to enter an MMP Connection URL, which you can find below. Please be sure to replace {app_store_id} with the iTunes app store ID of the relevant app: https://skad.api.kochava.com/v1/partner/facebook/model/apple-app-id/{app_store_id}

NOTE: Facebook optimizes for the day of the install and as a result you may see errors in the Facebook dashboard if you select anything greater than D0 in your conversion mapping.

NOTE: For more information on sharing your conversion model/conversion value schema with Facebook, refer to our support documenation.


Tracker Creation

There are no changes to the tracker creation process for SKAdNetwork tracking purposes. For more information on creating a Tracker, refer to our Create an Install Campaign support document.

NOTE: When creating a tracker for a media partner who has been integrated through Kochava’s SKAdNetwork certification program, you may notice their iOS integration templates contain SKAdNetwork-specific parameters.


Reporting

Standard Reports:

Two new standard reports have been added, called SKAdNetwork Install Summary and SKAdNetwork Conversion Value Summary. For more information, please refer to our Reporting support documentation.

 

Query:

A new table has been added in Kochava Query, called skad_conversions.