Identifier Data Import
Written by Dana Kang,Airbridge uses identifiers (IDs) to determine whether a user has performed an event for the first time. Import device IDs and user IDs to Airbridge and optimize campaigns based on accurate attribution results from the beginning of your Airbridge journey.
For more information about device IDs and user IDs, refer to this article.
Right after onboarding Airbridge, you may observe an increase in the counts of “first events,” such as first sign-ins and first purchases. This is due to the lack of user data that indicates whether the user has a history of performing an event.
To minimize initial data discrepancies, you can import device IDs and user IDs, a process also known as data backfilling.
Device IDs and user IDs must be imported for each individual event. Refer to the scenarios below.
Note
Airbridge can start collecting data once the user has updated the app to the Airbridge SDK-integrated version. This implies that the count of first events may be affected only after existing users update the app.
By importing IDs collected with the app install event, you can distinguish between users who have already installed the app and users who are installing the app for the first time. However, you can’t determine whether the user is performing any other event for the first time.
Let’s say a user who had already installed the app just updated to the Airbridge SDK-integrated version. When the user launches the app for the first time after the update, event counts will differ depending on whether the user’s ID has been imported to Airbridge. See the table below.
ID Upload |
Event Collected |
Impact on App Install Event Count |
|---|---|---|
ID not uploaded | When the user launches the app for the first time after the update, the first app open event is collected because no user data is readily available for Airbridge. | The first app open event is considered the app install event. Therefore, the count of the app install event may increase for a while. |
ID uploaded | When the user launches the app for the first time after the update, an app open event that is not a first event is collected because user data is readily available for Airbridge. | An app open event that is not a first event is simply an app open event. Therefore, the count of the app install event is not affected. |
By importing IDs collected with the order complete event, you can distinguish between users who have already made a purchase and users who are making a purchase for the first time. However, you can’t determine whether the user is performing any other event for the first time.
Let’s say a user who had already made a purchase just updated the app to the Airbridge SDK-integrated version. When the user makes a purchase for the first time after the update, event counts will differ depending on whether the user’s ID has been imported to Airbridge. See the table below.
ID Upload |
Event Collected |
Impact on First Order Complete Event Count |
|---|---|---|
ID not uploaded | When the user makes a purchase for the first time after the update, the first order complete event is collected because no user data is readily available for Airbridge. | The first order complete event is collected even though the user has a history of making purchases. Therefore, the count of the first order complete event may increase for a while. |
ID uploaded | When the user makes a purchase for the first time after the update, an order complete event that is not a first event is collected because user data is readily available for Airbridge. | An order complete event that is not a first event is collected. Therefore, the count of the first order complete event is not affected. |
Airbridge primarily uses device IDs and user IDs to determine whether a user has performed an event for the first time.
To see whether an event has been performed for the first time, use the “Is First Event per Device ID” or “Is First Event per User ID” GroupBy. Refer to the table below.
GroupBy | ID Used | Event Type |
|---|---|---|
Is First Event per Device ID | Device ID | App events |
Is First Event per User ID | User ID | App events, web events |
Let’s say there is a user who had made a purchase earlier, and you imported the user’s device ID but not the user ID. After the user updates the app to the Airbridge SDK-integrated version and makes a purchase for the first time, the above-mentioned GroupBys would show the following results:
Is First Event per Device ID: Shows “false” because the purchase is not considered a first event with the user’s device ID available.
Is First Event per User ID: Shows “true” because the purchase is considered a first event without any user ID available.
Importing IDs per event
Device IDs and user IDs must be imported for each individual event. Importing the IDs of users who have completed an event does not affect whether any other event should be considered a first event.
Let’s say you imported the device ID of a user with a purchase history. After the user updates the app to the Airbridge SDK-integrated version and makes a purchase for the first time, an order complete event that is not a first event would be collected. However, the device ID would not affect the counts of the app install event, the sign-up event, or any other events.
Importing priority
It is advised to prioritize importing device IDs over user IDs. Device IDs have a more direct impact on your ad spend and postbacks.
When you send first events to ad channels via postbacks, the first events are determined using the device ID. Therefore, you must import device IDs to optimize your campaigns or assess your ad spend based on first events.
The device IDs shall be imported by using CSV files.
Enter the device IDs for each event into an individual source CSV file, following the rules below to avoid any errors. You may also refer to the sample CSV file:
|
Rules |
|---|---|
File naming | - Format: EventName_DeviceID.csv - Example: airbridge.user.signup_DeviceID.csv |
File organization | - Device IDs and user IDs should not be combined into a single file. Keep them separate. - Include all types of device IDs, such as GAID, IDFA, and IDFV, in a single sheet. - The file size can’t exceed 500 MB. - Multiple files can be imported at once. |
Data collection period | - It is advised that device IDs collected within the last 1 year be uploaded. |
Data format | - Format: Version 4 UUID - Example: 550e8400-e29b-41d4-a716-446655440000 |
Data entry | - Only use column A in Sheet 1. Don’t use other columns or sheets. - Enter data from cell A1. - Enter the device IDs but nothing else. |
Attention
Make your request at least 2 business days before the first roll-out of the Airbridge SDK-integrated version of your app.
Once the source CSV file is ready, pass it to your Airbridge CSM. Make your request at least 2 business days before the first roll-out of the Airbridge SDK-integrated version of your app as data import takes time to fully complete.
You can check the import status with your Airbridge CSM.
The process for importing user IDs is the same as that for importing device IDs, except for the rules for the source CSV file. Create the source CSV file following the rules below and pass it to your Airbridge CSM to request import.
|
Rules |
|---|---|
File naming | - Format: EventName_UserID.csv - Example: airbridge.user.signup_UserID.csv |
File organization | - Device IDs and user IDs should not be combined into a single file. Keep them separate. - The file size can’t exceed 500 MB. - Multiple files can be imported at once. |
Data collection period | - It is advised that user IDs collected within the last 1 year be uploaded. |
Data format | - Format: Do not begin or end with spaces or quotation marks. - Example: abcd1234 |
Data entry | - Only use column A in Sheet 1. Don’t use other columns or sheets. - Enter data from cell A1. - Enter the user IDs but nothing else. |
It is advised that device IDs and user IDs collected within last 1 year be uploaded. The data collection period should be long enough to determine whether an event is a first event.
However, you may adjust the data collection period depending on the nature of your service or your marketing campaigns. For instance, if you decide to consider users who have not made a purchase in over 6 months and users who have never made a purchase the same, you should import IDs of users who have made a purchase in the last 6 months instead of 1 year.
During the ID import process which takes 2 business days, it is advised to pause any campaigns that may lead to an extensive collection of new device IDs or user IDs.
The device IDs and user IDs newly collected during this time can’t be imported to Airbridge. This means that events performed by the newly acquired users may be determined as first events even when the ID import has been successfully completed.
このページは役に立ちましたか？