How does MXO merge TIDs?

When an anonymous customer interacts with your brand for the first time, MXO creates a unique TID for that customer. Typically, customers interact across multiple channels. This means an anonymous customer may end up with multiple TIDs generated on different channels, across different browsers and on different devices. When MXO recognizes two TIDs as belonging to the same customer, it merges all activity for those TIDs into a single, primary, TID. The primary TID is the very first TID assigned by MXO to that customer.

For example, if a customer is browsing on your brands website and then contacts your call center, MXO track activity under two separate TIDs. When the customer logs into the website, providing recognition data that matches the CRM record, MXO merges all activity from the newer, secondary, TID into the older, primary, TID.

How does TID merging impact Data Export?

TID merges can happen anytime. This means that you may have a customer identified by one TID in your current export and by a different TID in your next export. You can view the TID merge events that cause this change in the merge_activity.csv file.

How do I merge all events for a customer across multiple Data Exports?

To see all events for a customer across multiple data exports, you need to link the event tables for those exports together. To do this, you need a primary to secondary TID map. This TID map will require iterative changes until all secondary TIDs link directly to a primary TID.

After the mapping is complete, you can join the event tables together. The process for completing this stage depends on the platform your brand is using:

• If using SQL Server, update the existing event TIDs using the TID map created above, before appending new events to the existing events table.
• If using SQL Data Warehouse, CTAS (Create Table As Select), amending TIDs in the older event table and then union them into a new table.

Merging Notes

• merge_activity.csv contains merges only for TIDs contained in the accompanying events file. For this reason, there should not be a time gap between the two event tables you want to merge. Any time overlap between event tables can be dealt with during the merge to prevent duplicates.
• You can merge multiple exports at the same time if all dates are covered.
• You do not need to maintain a master secondary/primary TID map. All required merges are available in each export. You may wish to maintain a master list for other reasons.
• Data Exports use a double quotation mark (“) as a text qualifier in .csv files. This qualifier is rarely used and may not be picked up by an automated assessment of your import method.
• customer-attributes.csv contains a snapshot of data, taken at the time of export. You must delete the entire table and replace it with the latest version for each update.
• Dimension tables are not incremental updates but may include disabled features. Merge new tables with existing tables before removing any duplicates.

File in Detail