Third-party integration
The Third-party Integration module provides integration with a range of third-party ad platforms and can sync ad campaign and ad monetization data from these platforms to the AE system. This article focuses on the core features of each page in the Third-party Integration module and its basic concepts. To learn how to connect each platform, see the child pages of this page.
1. Third-party integration home page
The Third-party Integration home page shows the integration plans that are currently connected and their basic information, including the name of the integration plan, the integrated platform and interface, the current integration status, the plan creator, and the last modified time.
Note that the new version adds a limit on the number of third-party integration configurations. The number of third-party integration configurations that can exist in each project at the same time cannot exceed this limit. If you reach the limit, you can delete plans you no longer need to free up configurations.
You can click Create plan in the upper right corner to go to the plan selection page and select a currently supported data interface to connect. You can also click the edit button on a card to go to the plan edit page and change the configuration of the integration plan:
2. Plan selection page
The plan selection page shows basic information about all currently supported integration methods. Select the integration method you need and click its card to go to its edit page.
The plan selection page also provides search, so you can search by platform name for the platform whose data you want to connect.
3. Configuration page
Clicking the edit button on the Third-party Integration home page, or clicking an integration method card on the plan selection page, takes you to the third-party integration configuration page. The image above shows the configuration page of the Adjust real-time callback plan. Because each platform has different data logic, the configuration pages of the platforms also differ.
The configuration page shows basic information about the selected plan, from which you can learn the data types the plan supports. You can also click the View Documentation link to open the integration plan documentation for the plan.
In addition, you can adjust the data integration rules on the configuration page. As shown in the image above, you can change the user identification field used when third-party data is stored, and whether events and user properties are stored. For deeper adjustments, edit the JSON in the Configuration section to control the specific configuration for pulling data. To learn how to configure each platform, see the integration plan documentation for the corresponding platform interface.
4. Data processing logic
The AE system now supports more than 40 third-party platform data integration plans, and the data provided by different plans differs in both structure and meaning. This section mainly describes the data interface categories and data types of third-party platforms, and the standardization services the AE system provides to smooth out the differences between platforms.
4.1 Interface categories
The features that third-party platforms use to provide data are called data interfaces. Data interfaces can be categorized along two dimensions: data granularity and integration type.
Data granularity
- User level: each record is one impression, click, or install by one user. The AE system converts this data directly into AE events or user data
- Aggregated metrics: each record contains analysis dimensions and various metrics but no user information. The AE system records this data as events on a virtual user
Take ad impressions as an example to understand the difference between the two data granularities:
- User-level ad impression data records information about the user and the ad. Each record represents one impression, so the total amount of impression data (that is, the number of events) is the number of impressions
- Aggregated metrics have only one or a few records, and the number of ad impressions is just one metric field in each record. Besides ad impressions, such a record usually also includes other metrics such as ad clicks, conversions, and cost, as well as ad dimension fields such as campaign and ad group. Each record represents the metrics at the finest analysis dimension, such as the clicks, impression rate, and total cost of one ad in a campaign
In general, user-level data can be used in all analysis models like other events, while aggregated metric data is not behavior generated by real users, so it can only be used in the event model, the retention model (mainly for calculating ROI), and SQL IDE.
Integration type
- Callback: the third-party platform sends its raw data back to the AE system in near real time, and data flows in continuously. All callback-type integration plans that the AE system currently supports provide user-level data. This is sometimes also called a Push API
- API: a data pull interface provided by the third-party platform. The AE system sends data pull requests to the third-party platform according to the configuration, and data comes in only after a pull request succeeds. Most API-type integration plans that the AE system currently supports provide aggregated metrics, and a few provide user-level data. This is sometimes also called a Pull API
The integration type affects how real-time the data is: callback data is basically real time, while API data has some delay. The two integration types are also configured differently: callback data is mainly configured on the third-party platform, and in the AE system you only need to copy the callback URL; for API data, you configure the pull token, time range, pull frequency, metrics, dimensions, and other information in the AE system.
4.2 Data types
Although data differs between platforms, the most essential data for ad campaigns and monetization falls into the following six types. On the Third-party Integration home page, you can see the data types each platform supports:
| Data type | Description |
|---|---|
| Attribution | Ad source information of a user after attribution. Attribution data can only be user level. You can see the attribution in the user's install event and user properties |
| Cost | The cost of user acquisition, either as aggregated metrics or as user-level cost for each user |
| Revenue | Revenue from monetization ads, either as aggregated metrics or as user-level revenue from each monetization ad |
| Impressions | Impressions of user acquisition ads or monetization ads, either as aggregated metrics or as impression data for each ad |
| Clicks | Clicks on user acquisition ads or monetization ads, either as aggregated metrics or as click data for each ad |
| Conversions | Conversion behavior of non-organic users in the game or app. On most platforms, the conversion behavior is the user install. Either aggregated metrics or conversion events for each user |
4.3 Standardized fields
To make it easier for marketers to analyze ad data from different platforms together, the AE system provides field standardization. Following preset logic, the AE system writes the core fields of the received data into an object field te_ads_object, unifying core fields across platforms. Most platforms that the AE system can connect to provide field standardization automatically.
| Standardized field | Description |
|---|---|
| te_ads_object.ad_account_id | Ad account ID |
| te_ads_object.campaign_name | Campaign name |
| te_ads_object.campaign_id | Campaign ID |
| te_ads_object.ad_group_name | Ad group name, or the Unit name for monetization ads |
| te_ads_object.ad_group_id | Ad group ID, or the Unit ID for monetization ads |
| te_ads_object.ad_name | Ad name |
| te_ads_object.ad_id | Ad ID |
| te_ads_object.placement | Ad placement |
| te_ads_object.media_source | Media source or monetization channel |
| te_ads_object.app_id | App ID |
| te_ads_object.app_name | App name |
| te_ads_object.platform | Platform, such as Android or iOS |
| te_ads_object.country | Country or region code |
| te_ads_object.currency | Currency of the cost or revenue |
| te_ads_object.impressions | Impressions |
| te_ads_object.clicks | Clicks |
| te_ads_object.installs | Installs |
| te_ads_object.cost | User acquisition cost |
| te_ads_object.revenue | Monetization revenue |
| te_ads_object.agency | Agency or agency channel |
| te_ads_object.conversions | Conversions |
5. Migration of legacy plans
Before version 4.2, most third-party integration plans supported by the AE system were configured by ThinkingAI staff. These configurations are called legacy plans. After version 4.2 is released, we automatically create new configuration copies of legacy plans, which you can view in the Third-party Integration module. For callback-type plans, we complete all migration work directly. For API-type plans, we migrate only some configuration parameters; the plan itself is still unavailable, and its status is shown as Solution transfer.
To avoid affecting normal data ingestion, we take the following approach:
- After the upgrade to 4.2, API-type plans continue to pull data with the legacy plans, and plans in the Solution transfer status in the Third-party Integration module are not actually executed. Callback-type plans use the new plans in the Third-party Integration module directly, with the status Connected, which means they are connected normally
- Unless you have special needs, we recommend that you keep using the legacy plans for existing plans
- To bring a new plan in the Solution transfer status online, contact ThinkingAI staff