Overview: Data imports configuration
Once you establish a connection between the data center and a data source, you need to define what type of data gets imported and how. This may be, for example, what the marketing department wants you to track and extract as data from the specific system. You may want to extract data that relates to a specific interaction or data that provides information about specific attributes of contacts.
For the different types of data you import, you need to set different import configurations, that is, set different properties of the import. For each data import configuration, you need to write a separate data query, be it an SQL query or JSON template query. Once you select how to import data from the data source, the query template changes to reflect this.
The following sections describe in more details the query configuration options, that is, type of data to import and query settings.
Interactions represent activities contacts complete, for example “download whitepaper X”. Progress Sitefinity Digital Experience Cloud uses the sentence data model to collect data. The sentence is the data unit that holds the information of what user activity was completed on the website, by whom, and when. The data model consists of:
- Data source
The system, in which the contact completed the interaction.
The contact that interacted.
Describes the actual activity that comprises the interaction. In the example above, the predicate is “download”.
NOTE: The predicate you define during data import configuration becomes an interaction choice when you define conditions and rules for marketing metrics for analytics and reports. Depending on your scenario, the predicate need to be unique.
The entity, with which the contact interacted
Tells the system when the interaction actually occurs
- IP address
For details on the sentence data model and its elements, see For developers: Sentences.
When you aggregate contact data from different sources, it is necessary that this data contains information about the user identification in the respective systems. Visitors are mapped to contacts usually when they log in or after they enter information in Sitefinity CMS forms. With this type of data import, you aim at mapping the contacts from the current data source to contacts from another data source through their IDs, so that Progress Sitefinity Digital Experience Cloud identifies the visitors from different sources as one and the same contact and uses these details in the contact profile.
Thus, contact mapping describes how a contact is represented in different systems. For example: mydomain.com/Contact1 <-> CRM/ContactA
For more information, see For developers: SubjectMapping.
Contact metadata comprise any additional information about a contact, apart from contact ID. This includes email address, company, phone, location, and so on. Contact attributes are represented by subject metadata. For details, see For developers: SubjectMetadata.
Settings for query results
Since query results may represent a very large amount of data and information, in the Settings section you can define how query results are paged. Paging breaks the data into fixed-size chunks that make the results manageable and reduces the amount of information and the speed with which data moves between the server and the client at one time. This affects the speed, with which results are displayed.
SQL queries settings
You configure paging of the results by specifying how many results you can upload (or, import) per page and then setting where a results page starts and where it ends using the Take parameter name and the Last read parameter name fields. Make sure the paging parameters concur with the ones in the query you write.
You also need to specify the number of results per page, which defines how much data is imported in one chunk. By default, paging is split into chunks of 1,000 results.
JSON template settings
You configure paging of the results by specifying the number of results per page, thus defining how much data is imported in one chunk. By default, paging is split into chunks of 25,000 results.