Workfront Library can consume a CSV file with up to 100,000 rows of data. While a larger file can take longer to import, Workfront recommends importing one file—if possible—because Workfront Library only processes one metadata import at a time. If you need to import multiple files, you must wait until the first file has finished before you can start another import.
Before formatting your CSV file, you need to set up your metadata schema in Workfront Library to match the structure of your previous content tool. To learn how, see Set up taxonomy metadata for Workfront Library.
All information (metadata tags, column headers, etc.) in the CSV file must exactly match what is set up in Workfront Library. Having slight variations—including extra spaces, abbreviations, or punctuation—could result in duplicate information being added to the metadata schema or metadata not mapping correctly to the asset. However, the fields are not case sensitive, so you don’t need to worry about capitalization.
For example, if you entered the tag US in the CSV file but Workfront Library had the tag U.S. instead, then Workfront Library would create a new tag. Entering us in the CSV file would correctly map to the metadata tag US.
The CSV file must match the same structure as the metadata schema in Workfront Library. You must include a column for each metadata field that you want to apply to assets listed in the CSV file.
The CSV file must also include these column headers:
|This column includes the filepath of the asset's location in Workfront Library. If you move an asset to a different folder prior to importing the metadata, you need to update the filepath to the new location. If an asset is in the My Content area of Workfront Library, the filepath column must be left blank for the metadata to map properly to the asset.|
|This column includes the filename of the asset in Workfront Library. The filename must include the filetype, such as .pdf or .png. Workfront Library does allow you to have the same filename for different assets with no repercussions from importing metadata, however, having duplicate filenames with the same filepath could result in only one of these rows mapping.|
Although the keywords column header is not required, you must include this column if you want to map keywords to your assets. To learn about requirements for the keywords column, see Keyword hierarchy.
Any dates in the CSV file must be listed in MM/DD/YYYY format. Otherwise, the metadata won’t map to the asset in Workfront Library.
To indicate that you want multiple tags selected from a metadata field, you must add the Pipe character to separate the tags. Adding a space to either side of the Pipe character results in spaces being added to the metadata, which could result in new tags being added to the metadata fields and assets. For example, typing “Fall 2018|Fall 2019” in a cell would select both locations in a multi-select field.
If you enter multiple tags in a column that maps to a single-select field, only the first tag maps correctly to the asset. To learn what happens when you add multiple tags to a single-select field or a multi-select field that aren’t currently included as selections, see How information updates.
If you were previously using commas to separate multiple tags within a field, you can use the SUBSTITUTE function to quickly replace all commas with the Pipe character.
If you have a metadata field titled “keywords” within your taxonomy metadata in Workfront Library, it will not update and you need to change the name both in Workfront Library and in the column header. Otherwise, this column will add any additional keywords in the keyword metadata fields to the asset(s).
Workfront Library allows you to create hierarchical keywords. Unlike the taxonomy metadata, the keyword structure does not need to be set up prior to importing the CSV file. Instead, you can create the hierarchy by importing the file. To learn more about keywords in Workfront Library, see Overview of keyword metadata in Workfront Library.
Any keywords that you need to add to your assets must be in a column with the column header “keywords.” The text listed in this column automatically maps to the keyword metadata in Workfront Library. You must indicate the keyword hierarchy within the cell, separating each level with the Greater Than Sign character. For example, “Fashion campaign>Fall>2019” is 3 levels deep with “Fashion campaign” as the root parent keyword.
If an asset has multiple keyword tags, you must separate them with the pipe character. For example, “Fashion campaign>Fall>2019|Fashion campaign>Fall>2018” is tagged with 2 keywords that have the same parent keywords at the first and second levels.
Adding a space to either side of the Pipe character or Greater Than Sign character results in spaces being added to the metadata, which could result in new tags being added to the metadata fields and assets.
To learn what happens when an asset already has keywords selected in Workfront Library and you add different keyword in the CSV file, see How information updates.
If an asset already has metadata information in Workfront Library but the metadata listed in the CSV file is different, the metadata for the asset updates to match the information in the CSV file. The way this works varies for the different types of metadata fields in Workfront Library.
|Regardless of whether metadata exists for the asset already, the tag entered in the column of a single-select field is mapped to the asset listed and all previous selections are removed from that field. If there are multiple tags listed for a single-select metadata option, only the first tag maps to the asset. However, the additonal tags listed will be added as options if they weren't already options for that field.|
|Regardless of whether metadata exists for the asset already, the tag(s) entered in the column of a multi-select field is mapped to the asset listed and all previous selections are removed from the field. If there are tags listed that aren't already options for that field, they are added as selections.|
Any keywords entered in this column are added to the current keyword list for the asset. All keywords previously added to the asset remain.
Leaving a cell blank in a column does not affect an asset's current metadata selections. To remove metadata information from an asset, you must remove them manually.
If you add a column that doesn’t exactly match a metadata field in Workfront Library, nothing in that column will map to the assets listed. To make sure that everything maps correctly, it is very important to make sure the column names, metadata field selections, and all other information listed in each row exactly matches the taxonomy in Workfront Library’s Setup area. Neglecting to do so can result in creating duplicate metadata tags or metadata not mapping correctly to an asset.
To learn how to import the CSV file, see Use Workfront Library’s import tool to add metadata to content.