Customer Attributes Datasets FAQ

This content is only for clients with the one-click DATASETS option in the top navigation bar. If you click DATASETS and see Product Data and Customer Data listed as menu options, then this content doesn't apply to your account. Instead, see the FAQ for Customer Data.

General

Does an identifier need to remain on the page 100% of the time for data activation to work?

An identifier is tied directly to a Monetate ID. For this reason, it only needs to appear during one session across all sessions. If a site visitor logs in to your site and then returns three months later and the same Monetate ID is present, Monetate recognizes the visitor on the first pageview without an identifier on the page. The identifier is tied directly to a Monetate ID that is set within a cookie that exists across multiple Monetate sessions as long as it is not deleted.

Data onboarding and Customer Views work together to unify IDs at all touch points for your customers: Data onboarding connects an ID to offline datasets, and Customer View connects that ID to other Monetate IDs to extract more behavioral data.

Automatic SFTP Updates

Should I provide partial file updates via SFTP?

Yes, sending partial file updates via SFTP is highly recommended. Partial updates process faster with more frequent updates.

Can I add a column to a dataset via SFTP?

No, you add one or a few columns, or attributes, to a dataset using the Add Attribute wizard that you access from the Attributes tab of the dataset's details page.

If you have many columns to add to an existing dataset, use the Full Update option in the manual file upload wizard that you can access from UPLOAD DATA on the dataset's details page.

Once you add the column(s), regardless of which method you use, only then does the SFTP feature accept new files based on the new schema that includes the new column(s). All subsequent files sent via SFTP must contain the added column(s) because if they don't, they won't match the new schema, resulting in a file-processing error.

What happens if a dataset has an error?

If a dataset throws an error, you cannot upload updates to it until you clear the error in the platform. Each error is specific to a dataset.