From Learning to Action: E-commerce Analytics with Olist Data

JonathanGar — Tue, 05 Aug 2025 14:12:49 GMT

Hello!

📊 What better way to consolidate what you’ve learned than through a real project?
As part of my transition into the data analytics world, I developed a complete dashboard using Olist's dataset – a Brazilian e-commerce platform. The goal: apply my skills in SQL, Power BI, and storytelling in a project as realistic as possible.

🛠️ I loaded the data into a MySQL database on AWS using Python, and handled most of the transformations in SQL, leveraging its efficiency to keep Power Query as lightweight as possible.

I designed a star schema model with three fact tables (orders, order_items, payments) and supporting dimension tables, including two calculated ones for Brazilian states and product categories. All DAX measures are centralized in a dedicated table for easier maintenance.

💡 Key highlights:
🔹 DAX Measures: From basics like Total Revenue and Profits to advanced ones like YTD, PYTD, MoM, and others used in custom visuals.
🔹 Parameters: Two core parameters allow users to control how many categories are displayed per page in a custom chart.
🔹 Bookmarks: Enable switching between views (time-based and interchangeable visuals) without duplicating report pages.
🔹 BINs: Used to group customer age ranges and number of credit card installments for more segmented analysis.
🔹 External Tools: Leveraged Bravo to simplify and optimize the data model.

📄 The dashboard contains 6 main pages:
🔸 Summary: Business overview, revenue and profit trends, and top categories and regions by profitability.
🔸 Products: In-depth analysis of product performance, pricing, purchase frequency, and profitability by subcategory.
🔸 Logistics: Assessment of delivery performance and delays, segmented by state and category.
🔸 Customers: Customer profile by age and buying behavior, including preferred payment methods.
🔸 Sellers: Performance comparison between vendors and their historical evolution.
🔸 Reviews: Analysis of customer satisfaction, trends over time, and how delivery delays impact ratings.

📌 Assumptions:
✔ Olist earns 30% of the product price.
✔ The customer pays for both the product and shipping.
✔ Olist handles the logistics.
✔ Target: ≤ 5% of late deliveries.
✔ Ratings ≥ 4 are considered positive.

eyJrIjoiNTU4Nzg3NDUtMTg0Ni00ZjU0LTg4NTAtMmQ4ZGUxZWNjMDRmIiwidCI6ImQ5ZDE4ZGQzLWQwMTItNGFjNS04NWViLTM2Yzc5MzZkOWRlMCJ9

Re: From Learning to Action: E-commerce Analytics with Olist Data

kenn8665 — Tue, 05 Aug 2025 15:25:20 GMT

This is great! Did you start with a template or would you be willing to share your template for this dashboard? I think my team would benefit from seeing this data all in one package, like you are presenting.

Re: From Learning to Action: E-commerce Analytics with Olist Data

humacakmak — Tue, 07 Oct 2025 15:32:04 GMT

hello it looks fantastic!! can i get the pbix. for template??

Re: From Learning to Action: E-commerce Analytics with Olist Data

Nabha-Ahmed — Tue, 04 Nov 2025 07:28:40 GMT

Beautiful balance between visuals and insights. Great work

Re: From Learning to Action: E-commerce Analytics with Olist Data

XBAnalytics — Tue, 24 Feb 2026 09:46:30 GMT

We’ve seen organizations significantly improve reporting accuracy after implementing structured Data Analytics Consulting frameworks that eliminate silos and automate insights.

topic From Learning to Action: E-commerce Analytics with Olist Data in Data Stories Gallery

From Learning to Action: E-commerce Analytics with Olist Data

Re: From Learning to Action: E-commerce Analytics with Olist Data

Re: From Learning to Action: E-commerce Analytics with Olist Data

Re: From Learning to Action: E-commerce Analytics with Olist Data

Re: From Learning to Action: E-commerce Analytics with Olist Data