1. Weight the evaluation with emphasis on the first contact
With Ad Ebis, you can select from the following weightings (conversion models) for the evaluation of each contact timing.
Even distribution : distribute the rewards equally across all contacts
Emphasis on first contact : 80% weighting is given to the first contact, and the remaining 20% is evenly distributed among other contacts
Emphasis on last contact : 80% weighting is given to the direct effect, and the remaining 20% is evenly distributed among other contacts.
Emphasis on first/last contact: 40% for first/first contact, 40% for direct effect, and the remaining 20% is distributed evenly among other contacts
Custom ... You can freely distribute the results yourself
In this experiment, we thought that by using "first contact emphasis" for the evaluation of discovery ads, we could evaluate the contribution of discovery ad contact to conversion (indirect effect).
Reference: Reallocation CV model settings (Ad Ebis official website)
How to set it up
In the top right corner of the Ad Ebis home screen, select Settings/Management > Basic Settings > System Settings
Select Display Options and then Edit Reallocation Conversion Model (Captured Area)
Select the desired conversion model and click "OK"
Reference: Setting the coefficients for the reallocation CV calculation model (Ad Ebis formula)
2. Cross-device functionality
AD EBISU links access from the same user using cookie phone number databases information (hereinafter referred to as user ID), but if the device or browser changes, the assigned user ID also changes, and the link is broken. However, by using AD EBISU's "cross-device function," it is possible to link broken access in two ways.
Machine Learning
By analyzing the access data collected by "Ad Ebis" using machine learning, it is possible to predict whether the fragmented accesses are from the same person.
Username
By reflecting the "site member ID" etc. in the user name of the conversion attribute, conversions with the same user name but different user IDs are judged as accesses from the same person.
The user name must be obtained from the site by setting up a tag for acquiring conversion attributes. For detailed setup instructions, please see here .
Reference: Cross-device function (Official Ad Ebis website)
How to set it up
Select the matching rule from "Machine learning" or "User name"
In the top right corner of the Ad Ebis home screen, select Settings/Management > Basic Settings > System Settings
Select the measurement option and then select Edit Cross-Device Linking (Capture Location)
Select either "Machine Learning" or "User Name" and click "OK"
Reference: Cross-device feature specifications, screen confirmation method, and setting method (Ad Ebis official website)
