Relational tables enable efficient information storage for behavioural data not supported by the contact table. Although relational structures are more efficient, they often have increasing amounts of data, which can eventually lead to a slowing down of tasks that use this data. Furthermore, regulations and best practices dictate that contact databases should not indefinitely keep expired information. In this context, automating the removal of relational data allows you to configure rules that specify your expiration criteria.
In this article, we will use the removal of transactions 24 months after the last purchase as an example. It is a common case when using implied consents: we want to keep the transactions for 24 months (as proof of implied consent) and remove them after that period.
Step 1: Create the Task
Go to Project → Data Management. Under Relational Data, click Removal, then click Create. Select the relational table to use for the task and give a label to the task (a name to easily identify the task). Tasks are limited to one per table.
Step 2: Add One or Several Clauses
Add one or several clauses (rule to identify the expired data):
Once you have configured the clauses, click Compute to validate the number of records targeted. If you want to view the targeted entries, you can use the search in a relational table page to recreate the clauses and view the data.
Step 3: Activate the Task
When you are done configuring the task, click Save, then Activate:
The launch of the task is determined automatically by the platform. The task will be considered ready to launch one week after the last successful execution. Then, each day during less busy hours, the platform will handle the most urgent tasks. The job may not run exactly one week after the last run based on the platform's schedule, but the run will be delayed until the platform can do the processing.
Consult the History
At the bottom of the Removal page, you will find the history of purge task launches. Review these results regularly to validate that the tasks are running correctly and for the appropriate number of records. You can filter this list by the table in order to view the complete removal history for a specific table.