---
title: "Discussion on \"Rebuilding Salesforce Data Logic in Power BI: Objects, Fields, and Relationships\" | Hashnode"
description: "Discussion on \"Rebuilding Salesforce Data Logic in Power BI: Objects, Fields, and Relationships\". Rebuilding Salesforce data logic in Power BI starts after the connector succeeds. A native Salesforce report arrives with joins, filters, groupings, and calculated output already embedded in its repor"
url: "https://hashnode.com/posts/rebuilding-salesforce-data-logic-in-power-bi-objects-fields-and-relationships/6ad1d9a7fe71536fc1509022"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Rebuilding Salesforce Data Logic in Power BI: Objects, Fields, and Relationships](https://metrica.hashnode.dev/rebuilding-salesforce-data-logic-in-power-bi-objects-fields-and-relationships?utm_source=hashnode&utm_medium=feed)

Rebuilding Salesforce data logic in Power BI starts after the connector succeeds. A native Salesforce report arrives with joins, filters, groupings, and calculated output already embedded in its repor

Oct 16 · 12 min read · metrica.hashnode.dev  
Read the article

-   [#power-bi](https://hashnode.com/tag/power-bi)
-   [#salesforce](https://hashnode.com/tag/salesforce)
-   [#crm](https://hashnode.com/tag/crm)
-   [#connector](https://hashnode.com/tag/connector)

## Written by

[Metrica Software](https://hashnode.com/@metrica)

@metrica

  
[View profile](https://hashnode.com/@metrica)

Enterprise-grade analytics connectors that enable businesses to seamlessly export data from Salesforce and SAP into Power BI.

## Responses 1

Markdown supported · ⌘↵ to post

[StackShortlist](https://hashnode.com/@veristria)

[Oct 17](https://hashnode.com/posts/rebuilding-salesforce-data-logic-in-power-bi-objects-fields-and-relationships/6ad1d9a7fe71536fc1509022/comment/6ad311695f1ff855ae3ff7b3)

Grain is the right place to start. Joining Opportunity, OpportunityLineItem and OpportunityFieldHistory in one table and getting plausible but wrong totals is the classic version of this, and plausible is what makes it dangerous. The 2,000 row limit on the Reports connector is also a good reason on its own to go to Objects early. How do you handle record level security from Salesforce, rebuild it as RLS in the model or keep the dataset limited to people who can see everything?