FAQs: Mobile App Measurement

What is a mobile measurement partner (MMP)?

An MMP (Mobile Measurement Partner) is a company that helps measure the performance of advertising campaigns across multiple channels, platforms, and devices to support app growth. MMPs can measure inventory served in-app or served on other devices where the action is either an app download or any post-install app event. Apps typically measured are on Mobile, Tablet, or CTV devices. For In-app placements, this is typically the only way to measure. MMP can use both deterministic and probabilistic measurement methods.

Do I need an MMP?

Any marketer with an app can benefit from working with an MMP. MMPs can analyze user behavior throughout the entire journey, including acquisition, installation, in-app behavior, and re-engagement.

Which MMPs is Quantcast integrated with?

Quantcast is integrated with Kochava, Branch.io, and Appsflyer.

Why should I enable Quantcast in my MMP?

Enabling Quantcast in your MMP will ensure you can accurately measure and attribute your app's performance. This will allow you to track the impact of your Quantcast campaigns and make informed decisions for optimization.

What capabilities are available through the MMP integrations?

Accounts can gain visibility into campaign performance in the MMP by enabling attribution measurement for mobile campaigns, creating attribution links, and sharing postback events to target and optimize mobile campaigns.

My MMP is not integrated with Quantcast. Can I still measure and attribute app performance?

You cannot measure and attribute app performance with MMPs where Quantcast is not an integrated partner or ad network.

Do MMPs offer fraud detection and solutions for apps?

Yes, MMPs provide fraud detection and solutions for apps. If an app install or post-install app event is detected as fraud, it is passed to Quantcast, and Quantcast filters out fraud app events in the app install goal optimization model.

Are there any limitations to the MMPs that are integrated with Quantcast?

Quantcast does not support cost and revenue passback for the MMP integrations.

What is app attribution?

App attribution assigns credit to paid media, the brand’s owned channels, and organic sources that result in a consumer performing a desired action, like an app install or post-install event. Attribution is performed at the impression and click level.

Please explain how app attribution functions.

Mobile Measurement Partners (MMPs) use a software development kit (SDK) that tracks data from an app, such as installs, post-install actions, and uninstalls. MMPs use deterministic and probabilistic methods to match ad impressions and clicks to installs and post-install app actions. Advertising identifiers are used to perform attribution. These identifiers are mobile and are referred to as MAIDS, short for mobile device IDs; MAIDS help determine which media source or sources contribute to an app install or post-install app action. The MAID for Apple is called IDFA (short for Identifier for Advertisers), while Google calls their IDs GAID (Short for Google Ad ID). Both MAIDs types, IP addresses, operating systems, device manufacturers, and other device information to attribute results.

Is there any difference between app attribution for Android and iOS?

Both iOS and Android platforms provide app attribution solutions, but there are some fundamental differences between them. First, the tracking method that iOS uses for IDFAs differs from what Android uses for Google Advertising IDs (GAIDs even though both IDFAs and GAIDs are unique device identifiers. Additionally, iOS prioritizes user privacy more than Android. Apple's privacy features utilize the App Tracking Transparency (ATT) framework. This framework seeks user permission to access app-related and personal data for tracking purposes by app developers. This has made app attribution more challenging for marketers since the implementation of iOS 14; Apple has made these changes to its privacy policies and required app developers to seek user permission before collecting their IDFA. This change has significantly decreased the number of users who allow their IDFA to be tracked.

What is deterministic attribution?

Deterministic attribution is an attribution method that matches ad delivery to app installs or post-install app events using the same advertising ID. Attribution doesn’t rely on a statistical model. The advertising ID is a mobile advertising ID (MAID) or identifier for advertisers (IDFA), also called a device ID. Deterministic attribution is the primary attribution method used by MMPs. The device ID is recorded from the app install or post-install app events and matched to the device ID recorded on the impression or click event from an ad impression.

What is probabilistic attribution?

Probabilistic attribution is an attribution method that matches ad delivery to app installs or post-install app events using different advertising ID types. Attribution relies on a statistical model to link the advertising ID types to match ad delivery to an app install or post-install app event. Probabilistic attribution is the secondary attribution method used by MMPs.

What attribution methods are supported for Quantcast campaigns?

Deterministic and probabilistic attribution methods are supported. Not all MMPs use probabilistic attribution by default, so it is essential to set up the preferred attribution settings when setting up Quantcast in the MMP.

Which types of paid campaigns require deterministic attribution?

Deterministic attribution is required for an app or CTV campaign measuring app installs or post-install app events where the device ID is captured on the impression or click event.

Which types of campaigns require probabilistic attribution?

Probabilistic attribution is necessary for an app or CTV campaign measuring app installs or post-install app events where the device ID is not captured on the impression or click event and for a web campaign measuring app installs or post-install events.

Are there any differences between IDFA and Google Advertising ID?

When a user accepts an app's App Tracking Transparency prompt, the iOS device passes a unique IDFA identifier, allowing developers and marketers to track user activity for advertising purposes. IDFA has two key features: 1.) users can reset it anytime, and 2.) they can opt out of IDFA-powered remarketing by enabling LAT or rejecting the ATT prompt.

When a bid request for inventory comes from an Android device, it carries the AAID, a unique, resettable ID used for advertising purposes, just like the IDFA. For tablets that have multiple users, the ID may be unique for each user. Encryption, technical details, and all other features of IDFA also apply to the advertising ID.

What app attribution use cases do Mobile Measurement Partners (MMPs) support for Quantcast media?

Does Quantcast support deep links or universal links?

Yes, deep links and universal links are supported.

What channels are supported for MMP measurement?

Web, In-app, & CTV

Can I enable campaign budget optimization for mobile campaigns?

No, campaign budget optimization is not supported for app campaigns

Can I get reporting for app events in the Quantcast Platform?

Reporting for app events in the Quantcast Platform is not supported. The results columns on the campaigns table page will be blank for campaigns with an app install objective. Log in to your MMP dashboard to view app installs and app reporting.

Can Branch measure app attribution on a desktop web environment?

No, Branch doesn’t support app attribution on a desktop web environment. Impressions are recorded on the desktop. However, VTA (View through attribution) is only applicable for mobile as the device identifiers (AAID and IDFA) are mandatory for attribution. Any conversions recorded without the device identifiers are counted as Organic.

Can Kochava measure app attribution on a desktop web environment?

Yes, Kochava supports app attribution on a desktop web environment.

Can Appsflyer measure app attribution on a desktop web environment?

Yes, Kochava supports app attribution on a desktop web environment. Impressions are recorded on the desktop. Any conversions recorded without the device identifiers are counted as Organic.

iOS App Attribution

What is SKAdNetwork (SKAN)?

In addition to the ATT Framework which asks permission for tracking with an identifier, Apple has an additional framework called SKAdNetwork (SKAN), which also focuses on Apple’s privacy-first attribution framework for iOS. SKAN performs attribution on-device without user-level identifiers. It provides accurate high-level attribution data for organic and paid media campaigns while respecting user privacy via Apple’s ATT framework regardless of content status.

How does SKAdNetwork work?

SKAN involves coordination between three key players in the mobile advertising ecosystem with different roles.

1. Ad networks and DSPs need to register and provide their network IDs, signed ads, and verify postbacks. 

2. App publishers need to integrate the network IDs and display the ads provided by the network/DSP.

3. App advertisers are responsible for providing app install validation and updating the Conversion Value

With the coordination of the three players involved, when a user clicks or views an ad, the ad partner (Quantcast) passes attribution parameters to SKAdNetwork along with the click or impression. If the ad leads to a conversion (install), the SKAdNetwork notifies the ad network and reports campaign information (network, publisher, and campaign ID) devoid of device-level details via postbacks to the ad network at least 24 hours after installation. Depending on privacy thresholds, SKAdNetwork may also report a conversion value from 0 to 63, which the advertiser can set to give some estimates of the users’ quality.

Do accounts need SKAN iOS attribution?

Accounts with an iOS app and who run paid media campaigns outside Apple Search Ads need SKAN. Accounts not running any paid media or only running campaigns on Apple Seach Ads do not need SKAN.

Does Quantcast support IOS campaigns with SKAN attribution measurement?

Quantcast cannot currently offer support for iOS campaigns that use SKAN attribution measurement. This is mainly because the adoption rate of our SKAN network ID among app publishers is currently low. However, we at Quantcast are continuously working to better understand how app publishers can integrate our SKAN network ID to scale iOS campaigns and we remain committed to supporting iOS campaigns with SKAN attribution measurement. It is important to note that Quantcast is a registered SKAN ad network.

Can an account still track iOS attribution without SKAN?

Some level of IDFA-based attribution may still be available when iOS app users agree to track through the ATT framework. However, most iOS app users opt out of such tracking, regardless of the appeal of an app’s ATT opt-in process. Without SKAN, the ability to see this opted-out user group is entirely hidden, which results in a partial understanding of the impact of advertising campaigns.

What are some updates of SKAN 4.0?

SKAN 4.0 comes with several updates to enhance the visibility of iOS campaigns. These updates include support for web-to-app attribution, fine and coarse conversion values, fixed window delay intervals at the end of the day (0-2, 3-7, and 8-35), and a 4-digit source identifier.

What are conversion values?

Conversion values are integers that you can map to app events within the SKAN attribution framework. Fine conversion values are integers from 1-63, and coarse conversion values are 3 values - high, medium, and low. You can map conversion values to app events in your MMP or directly with Apple.

How do I map conversion values in my MMP?

Please refer to the MMP-specific documentation for mapping conversion values. The links for mapping conversion values are AppsFlyer, Branch, and Kochava.