WELCOME TO WEI'S PORTFOLIO

PowerBI Case

Intro:

YourFashion is a global jewelry brand selling products to different markets. You as a marketing data analyst need to prepare a product analysis report to the marketing management team and also help the local marketing team to understand their own market performance. You need to explain the sales performance across different markets, top-performing and bottom-performing products, sales trend and the main driver for revenue growth. Meanwhile, the local marketing team also wants to see the data in their local currency. The corporate currency is DKK.

Solution design and implementation:

Step1: Case Analysis (4W1H)

Reports and dashboards - Kanban board.png

STEP2: Design the Logical Model

Create a star schema to display the relations between dimensions and facts.

Step3: Create a bus matrix (Facts vs Dimensions)

Step4: Analyze the dashboard and provide key insights in a storytelling manner

Click and see the dashboard here: PowerBI Dashboard

Data Science Case

Intro:

You need to use data collected from twitter bot which tweet any project reaching some milestone to predict what projects are more likely to succeed.

Solution design and implementation:

Step1: Business Understanding

The goal of this project was to gain insight into what factors constitute to making a campaign successful or not. Such an analysis could be useful for investors of future crowdfunding campaigns.

STEP2: Exploratory Analysis

Summary statistics
Succesrate categorized by country
It can be found that the country that had the highest rate of success was the United States.

Success rate categorized by main category
Looking at the figure below, it can be concluded that projects in the category ‘Dance’ were the most likely to be successful whilst projects in the category of ‘Technology’ were least likely to be successful.

Relation between success rate and date
Here, the lighter the tile gets, the higher the success rate is. With regards to the different months, we can observe that there aren’t huge fluctuations. Seasonal effects are therefore not really a factor that influences the success rate. If we look at the years however, we can clearly see that from august 2014 and onwards, there was a decline in the success rate which only recovered a little bit at the end of 2016.