Preload attribution using Google Play Auto Install

At a glance: AppsFlyer uses Google Play Auto Install (PAI) to attribute installs to preload campaigns, where marketers contract with mobile carriers, and app discovery partners to preload their app at the time the device is first turned on by the user. Note: This attribution method differs from APK preinstall attribution.  

PAI preload attribution

In preload campaigns, Original Device Manufacturers (OEMs), mobile carriers, and app discovery partners preload apps on devices that get downloaded upon first device activation.

Preload campaigns allow advertisers to reach a potentially large audience without running online campaigns.

Google Play Auto Install (PAI) (1 of 3 preload attribution methods) gives marketers an accurate and privacy-preserving mechanism to measure the performance of preload campaigns. This method provides marketers with deterministic attribution results and simultaneously ensures that user privacy is preserved since attribution is performed via referrer matching with a Google Play referrer generated independently of the device and user identifiers.

AppsFlyer gives the highest priority to preload attribution. This means that if a user engages with other campaigns before opening the app for the first time, the preload campaign is still awarded the attribution. Other engagements are regarded as assisted installs.

Note: Agencies can't enable preload campaigns for their clients.

Flow

  1. A new device is activated for the first time and predefined apps are automatically installed on it from Google Play according to rules set by the preload partner.
  2. The user launches a preloaded app for the first time.
  3. The AppsFlyer SDK queries the Google Play referrer API.
  4. AppsFlyer attributes based on the data retrieved.

PAI-attribution-flow_en-us.png

Data availability

Google Play Auto Install (PAI) preload performance data is available in:

  • The Overview dashboard and in aggregate reports, aggregated at the preload partner (media source) and campaign levels.
  • Raw data reports.

Raw data includes the following fields:

  • Media_source = Partner ID
  • Match_type = preload_pai
  • Campaign_name = preload campaign by <preloadpartner ID>

Set up Google Play Auto Install (PAI) preload attribution

To set up PAI preload attribution:

  1. In AppsFlyer, from the side menu, select Collaborate > Active Integrations.
  2. Select the preload partner.
  3. Turn on Preload attribution
  4. Set the Preload lookback window duration. Range: 1-180 days.
  5. Click Save Integration
  6. Let the preload partner know that you are ready to attribute their campaigns.

Partners supporting Google Play Auto Install (PAI) preload campaigns

Partner PID

Xiaomi

xiaomipai_int

Transsion

shalltrypai_int

Oppo

oppopaipreinstall_int

Vivo

vivootapreload_int

TCL

tclpai_int

Motorola

motorola_int

MITV

mitvpai_int

ZTE

ztepai_int

Lenovotab

lenovotabpai4p_int

KDDI Pre-Install

kddipai_int

Postbacks for preload partners

As part of their integration with AppsFlyer, preload partners define the content and endpoints of postbacks sent to them by AppsFlyer.

When included in a postback, macros contain user-relevant data. Supported macros are listed in the following table:

  • Source: Indicates where the data originates from:
    • SDK: AppsFlyer SDK embedded in the app or server-to-server API.
    • AppsFlyer: After processing by AppsFlyer.
Postback macro for preload partners (parameter name) Source Description
app_name SDK App name set by the advertiser
country_code AppsFlyer Country Code using ISO 3166 (alpha-2) Example: US, CN.
install_time SDK Install timestamp
language SDK Language (locale) reported by the device and set by the device OS.
app_id SDK Unique app identifier in Appsflyer
app_version SDK App version set by the app developer in the app code
device_type SDK The commercial model name of the device. Example: SM-GM284 or iPhone7
os_version SDK Device operating system version

Postback example

http://YourCompanyDomain.com?pid=(preinstall_partner)&app_name=(app-name)&app_id=(app-id)&install_time=(install-unix-ts)