Attribution Models Report
Posted: Mon Dec 23, 2024 9:27 am
Content
New functionality
The new report will allow you to evaluate the impact of advertising channels at different points along the conversion path. This will help answer the question: "Which channel actually brings in targeted requests?" and plan the redistribution of your advertising budget even better.
New functionality
In your CoMagic personal account, in the Attribution Tools telegram database search group of reports, the Attribution Models report is available. This is a set of rules that determine the principle of value distribution among interaction points. By choosing one or another model, you can compare the effectiveness of advertising channels from different "angles of view" and use the one that is most suitable for your business.
The report introduces a new chart type that allows you to compare the influence of a channel across different attribution models.
In the column settings, you can add metrics for the desired attribution model.
Each indicator has its own symbol of belonging to the attribution model in the column settings, in filters, and also in the report download:
᎐᎐᎒ Last Interaction - This model identifies the last referral source for each conversion, without taking into account the visitor's browsing history.
᎐᎒᎐ Last Indirect Interaction - This model identifies the source that preceded the direct interaction source (where the conversion occurred) for each conversion.
᎒᎐᎐ First interaction - when using this model, the first source of the visitor's transition is determined for each conversion, i.e. the source that first brought your visitor to the site.
᎐᎐᎐ Linear - This model divides each conversion by the number of visits the visitor made before the conversion, meaning each visit is given an equal share of the conversion.
᎐᎓᎒ Taking into account the recency of the interaction — when using this model, a higher conversion value is assigned to those visits that are closer to the moment of conversion. The value is halved for every 7 days before the conversion. That is, a visit that was 7 days before the conversion has half the value of a visit on the day of conversion. At the same time, a visit that was 14 days before the conversion has 4 times less value.
᎓᎐᎓ Position-Based (40-20-40) - With this model, each first and last visit is assigned 40% of the conversion, and the remaining 20% of the conversion is evenly distributed among the remaining visits.
᎓᎒᎓ Position-Based (20-60-20) - With this model, each first and last visit is assigned 20% of the conversion, and the remaining 60% of the conversion is evenly distributed among the remaining visits.
Depending on the attribution model, the value of the session that precedes the request will change. At the same time, the number of sessions in the dimension remains unchanged for all attribution models.
The report is built with a retrospective window equal to 180 days, i.e. the report includes data on the visitor's sessions and their requests for the last 180 days. The retrospective window is adjusted using a special control.
The report is generated based on visitor ID.
The data in the report is broken down by site and depends on the user's access rights to the site and advertising campaigns.
New functionality
The new report will allow you to evaluate the impact of advertising channels at different points along the conversion path. This will help answer the question: "Which channel actually brings in targeted requests?" and plan the redistribution of your advertising budget even better.
New functionality
In your CoMagic personal account, in the Attribution Tools telegram database search group of reports, the Attribution Models report is available. This is a set of rules that determine the principle of value distribution among interaction points. By choosing one or another model, you can compare the effectiveness of advertising channels from different "angles of view" and use the one that is most suitable for your business.
The report introduces a new chart type that allows you to compare the influence of a channel across different attribution models.
In the column settings, you can add metrics for the desired attribution model.
Each indicator has its own symbol of belonging to the attribution model in the column settings, in filters, and also in the report download:
᎐᎐᎒ Last Interaction - This model identifies the last referral source for each conversion, without taking into account the visitor's browsing history.
᎐᎒᎐ Last Indirect Interaction - This model identifies the source that preceded the direct interaction source (where the conversion occurred) for each conversion.
᎒᎐᎐ First interaction - when using this model, the first source of the visitor's transition is determined for each conversion, i.e. the source that first brought your visitor to the site.
᎐᎐᎐ Linear - This model divides each conversion by the number of visits the visitor made before the conversion, meaning each visit is given an equal share of the conversion.
᎐᎓᎒ Taking into account the recency of the interaction — when using this model, a higher conversion value is assigned to those visits that are closer to the moment of conversion. The value is halved for every 7 days before the conversion. That is, a visit that was 7 days before the conversion has half the value of a visit on the day of conversion. At the same time, a visit that was 14 days before the conversion has 4 times less value.
᎓᎐᎓ Position-Based (40-20-40) - With this model, each first and last visit is assigned 40% of the conversion, and the remaining 20% of the conversion is evenly distributed among the remaining visits.
᎓᎒᎓ Position-Based (20-60-20) - With this model, each first and last visit is assigned 20% of the conversion, and the remaining 60% of the conversion is evenly distributed among the remaining visits.
Depending on the attribution model, the value of the session that precedes the request will change. At the same time, the number of sessions in the dimension remains unchanged for all attribution models.
The report is built with a retrospective window equal to 180 days, i.e. the report includes data on the visitor's sessions and their requests for the last 180 days. The retrospective window is adjusted using a special control.
The report is generated based on visitor ID.
The data in the report is broken down by site and depends on the user's access rights to the site and advertising campaigns.