Data-Cleaning-and-Analysis-with-Python

Task

To evaluate whether agressive discounting is beneficial in the long-run.

Overview

Eniac's board noticed that the last quarter's increasing sales volume did not translate to higher revenue.

Marketing: Believes that discounts are beneficial in the long-run for customer acquisition and retention.

Board: Is concerned that their positioning as a premium quality provider is compromised by agressive discounts.

Context

Eniac is an online marketplace specializing in Apple-compatible accessories. The board wants an immediate answer to whether the company should continue discounting or not. The data is compromised, therefore before analysis we must clean the data and assess its quality, in addition to defining how it impairs our decision-making ability.

Challenge:

How to clean and assure the data's quality without losing too much information, so as to retain enough data for decision making?

Approach

Evaluate the database:

1. Clean data for unreadable entries, duplicates and other obvious errors
2. Assess the remaining data for quality - remove compromised orders
3. Note the constraints the loss of data caused in our ability to make decisions
4. Basis of decision making: comparison between the recommended prices, and product catalog and the actual sales
5. Note recommendations; how to improve data collection and further research questions

Deliverables

5 minute PowerPoint presentation found here to the Board of Directors, that summarizes the findings and suggests a course of action. Python code is found here.

Colab Files

1. Files starting with 2 are the data cleaning files, each table its own file
2. Files starting with 3 are the data quality files
3. Files starting with 4 are the data analysis files

Skills & Tools

1. Data Cleaning & Quality Assurance
2. Data Visualization & Storytelling
3. Colab & Jupyter Notebook
4. Python: Pandas, Seaborn, Matplotlib