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

Community

Connector Improvement: Shopify - Requesting bundle-related order details

Answered
Varinderjit Kaur User

Currently, order-level line items do not provide a reliable way to determine whether they belong to a bundle nor do they indicate which bundle product each component is associated with. This makes identification of bundle vs non bundle orders really difficult.

We can use combination of product_variant, bundle_component, bundle_component_option_selection and bundle_component_variable, they can only be used as a partial workaround. This approach introduces blind spots, especially when bundle definitions change frequently and the data is not refreshed in real time. As a result, we risk misclassifying orders and losing end-to-end traceability of bundle composition. There are fields available via shopifyql to determine the bundle_product_id, line_item_is_bundle etc for better order level visibility, Attaching snips for reference.

Please let us know when are you planning to include bundle related details in orders.

 

Please sign in to leave a comment.

Comments

8 comments

Date Votes
  • Official comment
    Sandeep Preetam Saila Fivetranner

    Hi Varinderjit,

    Thank you for the feature request and for sharing the detailed context around bundle identification challenges.

    Today, we model bundles as products and rely on the relationship
    ORDER → PRODUCT → BUNDLE_COMPONENT to infer bundle composition. This helps us keep the schema normalized and avoid denormalizing bundle-specific fields directly into the ORDER or ORDER_LINE_ITEM tables.

    To better understand the gap you’re seeing, it would be helpful to clarify one point:

    • In your setup, why does the ORDER → PRODUCT → BUNDLE_COMPONENT mapping not give accurate results? Is this primarily due to frequent changes in bundle definitions? If yes - can you elaborate with an example?

    Understanding where this relationship breaks down will help us evaluate whether syncing bundle-specific attributes (like bundle_product_id or line_item_is_bundle) at the order level is necessary, or whether there’s a more reliable way to preserve traceability without denormalizing the data.

    Happy to discuss this over a meeting. You can book time with me here.

    Best,  
    Sandeep
    Product Manager

  • Varinderjit Kaur User

    Hi Fivetran team, Following up on this. Thanks!

  • Varinderjit Kaur User
    • Edited

    Hi Sandeep,

    Thank you for the response.

    I understand the concern regarding normalization. However, relying on a normalized path (Product → Bundle Component) is exactly what is causing this break for historical orders. Normalization works for static relationships but our bundle compositions are dynamic. Once a bundle definition changes, we lose the historical link.

    Example (Order 6209789329565):

    • At the time of this order, the component 'Holiday Celebrations Pellet Blend' (8784397631645) was part of the 'DualFire 1200 Smokebox Bundle' (8557811466397).

    • Today, that same component is associated with four different bundles (e.g., 8784849535133, 8679401259165, 8621039747229, 8682033479837).

    • The Result: If we try to 'normalize' this by looking up the current bundle for that component, our reports will incorrectly show four bundles instead of the one the customer actually purchased.

    While normalization is the goal, applying it to transactional data can break our historical snapshots. To lock the record at the time of sale, we need the bundle_product_id. Otherwise, as bundle definitions evolve, our historical reporting will shift which compromises our data integrity.

    If you need more details, Please let me know if you are available tomorrow(20/01) at 3 PM IST

    Best,

    Varinder

     
  • Varinderjit Kaur User

    Hi Sandeep, Any update on this?

  • Varinderjit Kaur User

    Hi Sandeep, Can please you share details here?

  • Sandeep Preetam Saila Fivetranner

    Hi Varinderjit,

    We will pick this up for analysis in the current sprint and brainstorm potential solutions. We will update the portal with the way forward by the end of the sprint, February 18.

    Best regards,
    Sandeep

  • Varinderjit Kaur User

    Hi Sandeep, Any updates on this?
    We are revisiting this again, Your help would be highly appreciated.

  • Varinderjit Kaur User

    Hi Sandeep, Following up on this.