{"id":1054,"date":"2025-12-22T22:19:24","date_gmt":"2025-12-23T03:19:24","guid":{"rendered":"https:\/\/blog.data-principles.com\/?p=1054"},"modified":"2026-01-06T18:13:23","modified_gmt":"2026-01-06T23:13:23","slug":"start-your-bi-journey-with-the-data-model","status":"publish","type":"post","link":"https:\/\/blog.data-principles.com\/index.php\/2025\/12\/22\/start-your-bi-journey-with-the-data-model\/","title":{"rendered":"Start Your BI Journey with the Data Model"},"content":{"rendered":"\n

By Jeremiah Willett<\/em><\/p>\n\n\n\n

\n
\n

Data Principles is currently building a training course for a client: Introduction to Power BI. One of the most important topics in the course is data modeling in Power BI. Many introductory training courses to Power BI start with building the visuals or Data Analysis Expression (DAX) tricks. However, at Data Principles, we believe that building visualizations in a BI tool (Qlik, Power BI, Tableau, etc.) should be the final step in the data storytelling process. First, and most importantly, we tell a story with our data model. <\/p>\n\n\n\n

The problem with starting with the drag-and-drop visuals and DAX tricks when teaching Power BI to a student is that there will be no foundational understanding of the data and how each data element relates to another. This isn\u2019t just a problem of not knowing something; building visuals without a complete data model will ultimately lead to the dashboards breaking and wasting time. <\/p>\n\n\n\n

An example case that I used in the training is a political campaign analytics dataset. We had various tables with campaign-related data: staff activity, donations, events, and voter outreach activity. Each of these campaign elements interacts with the other, and that is the story <\/p>\n<\/div>\n\n\n\n

\n

that we want to tell in our data model. <\/p>\n\n\n\n

Every dataset has a narrative structure. Dimensions are the nouns of our story: Campaigns, Staff, Donors, Dates, Events. Facts are the verbs or actions: Donations, Volunteer Hours, Doors Knocked, Event Attendance, Donor Phone Calls. When the nouns and verbs in the dataset are organized properly, the business questions will flow naturally: <\/p>\n\n\n\n