<![CDATA[This research examines the implementation of Business Intelligence (BI) for the creation of an Indonesian e-commerce buyer dashboard in 2024 with the aim of increasing the visibility of operational KPI and demonstrating a reproducible pipeline from data cleaning to visualization. The main issues addressed are the quality of order-level data (provincial writing variants, date format, numerical values, and PII anonymization) as well as the need to calculate buyer metrics (unique buyers, repeat rate) which is rarely ava1ilable in public aggregate data. The methods used include: (1) data cleaning and harmonizing using OpenRefine; (2) numerical transformation and validation with Python (pandas); (3) creating interactive worksheets and dashboards in Tableau (sales map per province; monthly trend line; bar with avg sales per product; sales pie by gender); and (4) sensitivity analysis to assess the impact of cleaning step variation on buyer-level KPI. Using the order-level dataset of cleaning results (1,000 transactions), a total revenue of Rp 2,298,975,000, 1,000 orders, and 178 unique buyers were found; seasonal patterns were seen with a peak in the fourth quarter and revenue concentration in urban areas (DKI Jakarta, West Java). The top-10 products contribute a significant portion of revenue, and repeat buyers show an important role in the sales structure]]>
Copyrights © 2025
