Skip to main content
Docs
Feature Requests Onboarding Trust Center Status Page

Community

Connector Improvement: Match Salesforce naming conventions for tables and columns

Completed
Brett Bevers User

My data engineering team manages the data platform for several internal customers. We are using Fivetran's Salesforce connector to ingest one customer's data into our Databricks lakehouse. In this case, we are landing bronze tables and the customer intends to handle data modeling from that point. We are disappointed that the naming conventions for tables and columns landed by Fivetran differ from what is seen in Salesforce. In particular, the conversion from camelcase to snake-case makes it difficult for analysts to switch contexts between Salesforce and our lakehouse. 

We want the option to avoid removing or introducing characters in table or column names compared to what is seen in Salesforce.

Please sign in to leave a comment.

Comments

3 comments

Date Votes
  • Official comment
    Egidio Terra Fivetranner

    Hi Brett,

    Excited to share a new feature to address Fivetran naming normalization: Source Naming. It preserves Salesforce table and column names directly from the source system and will prevent collisions.

    You can learn more in the new documentation section here: https://fivetran.com/docs/core-concepts/naming-conventions#namingconventions 

    Best,

  • Egidio Terra Fivetranner
    • Edited

    Hi Brett, 

    Thanks for sharing your feedback! We completely understand the issue and are already working on ensuring that the Salesforce field names appear exactly the same in the destination. This update is planned to be released soon. We'll keep you posted.

    Best,

     

     

  • Antonio Ruby Barreto User

    Adding to the reasons for being able to opt-out the normalization, the automatic case conversion per Fivetran creates conflicts in column names because 2 columns of Salesforce end up being normalized to the same name in the lakehouse (e.g. ManualBuild__c and Manual_Build__c both go to manual_build__c).

    The team configuring Salesforce sometimes creates similar columns like this during migration efforts, and then it becomes complicated in the lakehouse to manage the migration if we can't receive both columns together.