E-commerce Analytics Project

Overview

This was an add-on final project from the Data Analyst bootcamp at MySkill, on top of the regular coursework. Nine of us split into three tracks, SQL, Python, and data visualization, to work through a stakeholder case using a dummy Tokopedia-style e-commerce dataset from 2022. I was on the data visualization track with two teammates, Aziza and Diah Natalia, building a two-page dashboard in Looker Studio for a fictional marketing team.

The Build

The Full Deck

What I Built

Since I already knew Tableau and Power BI from other projects, I ended up initiating the dashboard design early while the rest of the team was still working through the case. Before touching Looker Studio, I cleaned the dataset in Excel: checked for duplicate rows, none, and blank cells, none, then built three calculated fields the dataset didn't have yet: Net Profit (sales before discount minus cost of goods sold), Value Transaction (splitting valid and invalid payments), and AOV (sales divided by unique orders). Page one tracked Value Sales, Net Profit, and AOV over 2022 with a combo chart and four slicers for date, category, payment, and transaction validity. Page two dug into product and customer performance, plus a specific question from the brief: check whether the Mobile and Tablet category had any transactions paid through Jazzwallet. I also helped put the final PPT together since the deliverable had to be submitted as a presentation.

What I Learned

This was the first time I had to defend a dashboard against a written case brief instead of just building whatever looked good. Every chart had to answer a specific stakeholder question, so I kept rereading the brief instead of guessing what the marketing team wanted. The Jazzwallet question ended up being the most interesting part. My first instinct was to explain a low adoption number, but the actual answer was zero, not a single transaction all year, which changed the story from weak adoption to the payment option not reaching that segment at all.

Outcome

The dashboard answered both halves of the brief. Page one showed a clear seasonal pattern: Value Sales, Net Profit, and AOV all spiked together in April around Ramadan, Value Sales and AOV spiked again in August and September without Net Profit following, which pointed to heavy discounting during the 8.8 and 9.9 sale events, and all three dropped through Q4. Page two confirmed Mobile and Tablet as the biggest revenue category by far, around 8.7 billion in sales from 1,200-plus units, and answered the Jazzwallet question directly: zero transactions in that category used it all year, which we flagged as a real gap instead of a coincidence.

Tools & Stack