The purpose of this project is to analyze and interpret a dataset with information about the customers of a store. We don't know what type of business it is except that they sell grocery items. We only know what types of products they provide, the ways in which customers have previously purchased and some metrics about marketing campaigns launched.
It is up to us to determine what type of clientele we have, what their relationship is with the products on sale and to interpret how the campaigns launched have gone. We will also try, with the final reading of all the data, to give some guidelines to the marketing department to increase sales.
- Data Cleaning: We commence by meticulously cleaning the dataset, addressing missing values, outliers, and inconsistencies to ensure the integrity and reliability of our subsequent analyses.
- Data Wrangling: After data cleaning, we've undertaken data wrangling tasks to adapt the dataset for improved categorical interpretation of customer behavior and gain better insights into continuous variables related to their activities.
- Exploratory Data Analysis (EDA): During EDA, we have not only identified various customer profiles but also detected purchasing patterns within these profiles, unraveling insights, patterns, and correlations crucial for feature selection and model development.
- Buyer Persona: By employing customer segmentation techniques and analyzing learned patterns, we craft comprehensive profiles of our most frequent customers.
This project is developed through three notebooks.
- INITAL SET UP: In this first Nb we performed the tasks of data extraction, standardization of variables for the subsequent step, to perform the EDA.
- EDA: This Nb is intended for the interpretation of data from the previously cleaned dataset. We break down the statistics of the customers, their consumption patterns and ways of purchasing and their activities in campaigns.
- BUYER PERSONA DESIGN: In this final step, we use the segmentation defined during the EDA to create detailed profiles of our most frequent customers and clearly visualize their consumption habits and purchasing trends.
As a final presentation of what has been developed and obtained in this project, we also include some explanatory slides. You can download the slides from here.
Five days, from 02/25/24 to 03/1/24
With this work, I intend to show a little of the skills acquired during the Ironhack Data Analyst bootcam. I have focused on using EDA to try to get the most out of a base dataset on purchase statistics and customer actions in a store in order to define several suggestions for improvements for the Marketing department.
I had one week to complete this project. The first day I also thought about including ML classification model to predict if the next customers to be included in our database were going to accept the next marketing campaign, but due to lack of time I finally decided to focus on analyzing in depth the customer segmentation and develop buyer persona profiles of the most frequent customers.
I hope you find the above interesting. I have tried to mix in this project my new vocation as Data Analyst with my previous background, Digital Marketing Strategist.
Best regards and see you in other Gits! 