The Attribute and Recommendation Syndication is used to export user attributes, recommendations and/or predictive scores to either the S3 bucket that we provide or a location outside of Blueshift.
If you want to store the files on an Amazon S3 or SFTP location of your choice, ensure that you set up those stores using the following documents:
- On the left navigation menu, click Syndications to go to the Syndications index page.
- On the top right corner, click Syndication + and then click Attribute and Recommendation (if you see a dropdown).
- On the page that appears, specify a name for the syndication.
This is used purely for organization purposes and must be unique for this syndication. Tags field provides options to categorize and organize the syndication. For more information, see Tagging.
On the page where you specify the Syndication's settings, the configuration is organized in six steps:
The first step is the configuration stage.
The Segment field provides the option to select the segment that you want to syndicate/export out of Blueshift.
If control groups are enabled for your account, you can choose whether this syndication should honor it or not. For more information, see Control groups.
Filename Prefix will be appended to the name of the file that is ultimately uploaded. This is required for organizational purposes.
Send Completion Email To is an optional field. Enter a single email or multiple emails (separated by commas) to be emailed once the syndication finishes calculating and uploading the file(s).
Click Continue after you configure your syndication.
After configuring the actual file that will be created, you now have to choose where the file will ultimately be uploaded.
Click on the Blueshift Bucket to upload this file to your Blueshift Bucket - the customer bucket where all other files are stored. Or, you can choose the S3 or SFTP location that you added in Blueshift.
Click Continue after you choose a location.
Specify the schedule of the syndication.
Start Time specifies when this syndication will launch.
End Time specifies when this syndication will stop.
Execution schedule is how often this syndication will run and is further elaborated in the following table:
|One-time||This setting will export the segment on a one-time basis.|
|Daily||This setting will export the segment daily, at the Start Time specified above.|
|Weekly||This setting will export the segment weekly, on the Start Time and 7 days afterwards, and so forth.|
|Custom||This setting will export the segment every N number days, where N is a number that the user inputs.|
Click Continue after you specify the schedule.
The next step is adding user fields for this syndication. These are the user level attributes that will be outputted to the CSV file. For good practice, you should probably choose at least one unique identifier for each user. e.g. email. Otherwise the outputted information may not be usable.
To add more fields, click the + ADD USER FIELD button and add more user fields. You can add as many fields as you need.
Click Continue after you specify the user fields.
If recommendations are enabled on your account, you can add recommendations to your outputted file.
This is an optional step.
You need to select the recommendation algorithm (either a predefined algorithm or any created via our Recommendation Studio), total the number of products you want returned (up to 20) and the different product fields that you desire.
The column names are predefined and will be generated from the product field name and the number of products.
If Predictive Scores are enabled, you can add predictive scores to your outputted file by clicking on the following:
Adding the predictive scores is similar to adding user fields. Select the score and enter the name of this column.
After all of the appropriate settings have been filled out, save this syndication then hit launch. This will activate this syndication. Attributes and recommendations will be computed and uploaded to the selected destination.
There are three pieces of information that are relevant per destination:
- Users in Segment: The total number of users that qualify for the conditions defined in a particular segment.
- Last Successful Run: The last time this Attribute and Recommendation syndication ran successfully, from calculating the users to uploading the final file.
- Next Scheduled Run: The next time this Attribute and Recommendation syndication will run.