COFINFAD: Colombian Fintech Financial Analysis

A visual analytics project examining how socio-demographic and behavioural factors shape customer financial activity, digital engagement, and satisfaction in the Colombian fintech ecosystem.

View Project Poster

Project Overview

This project examines customer transaction behaviour using the COFINFAD dataset. The study focuses on understanding how socio-demographic and behavioural factors shape financial activity and customer outcomes within a fintech environment.

The project adopts a multi-perspective analytical approach, combining:

  • Exploratory Analysis
  • Customer Segmentation
  • Predictive Modelling

Project Objective

The objective of this project is to understand the underlying behaviours behind customer transactions and evaluate how socio-demographic factors such as age, income, education, and household characteristics affect financial activity.

Through visual analytics and statistical techniques, the project aims to:

  • identify meaningful customer patterns
  • detect behavioural differences across demographic groups
  • generate insights to better interpret fintech user behaviour

Project Components

This project is structured into three key analytical modules:

1. Exploratory Analysis

Investigates relationships between socio-demographic characteristics and transaction behaviour using descriptive statistics, visualisations, and confirmatory analysis.

2. Clustering Analysis

Applies customer segmentation techniques to identify groups of users with similar financial and behavioural profiles, revealing distinct user segments.

3. Predictive Modelling

Develops models to predict customer satisfaction and transaction behaviour using demographic, behavioural, and engagement variables.

Dataset

The analysis is based on the COFINFAD dataset, which contains both customer-level and transaction-level information from a Colombian fintech environment.

The dataset includes:

  • socio-demographic attributes
  • behavioural and engagement variables
  • transaction frequency and value metrics
  • customer satisfaction indicators
  • financial activity features

Dataset Source:
COFINFAD Dataset on Mendeley

Website Structure

This website documents the development of the project across multiple stages:

  • Proposal
  • Methodology
  • Findings
  • Poster
  • Meeting Minutes
  • Team
  • Shiny Application

Findings

The COFINFAD project brings together confirmatory analysis, clustering, and predictive modelling into one integrated interpretation of customer behaviour in the Colombian fintech ecosystem.

The final findings highlight how demographic variables provide useful context, while behavioural patterns, engagement signals, and transaction activity offer stronger explanatory value across customer groups.

The full interpretation is presented in the Findings page, where results from all analytical modules are consolidated into a single narrative.

Team Contribution Overview

Each team member leads a core analytical component:

  • Exploratory Analysis: socio-demographic relationships and transaction behaviour
  • Clustering Analysis: customer segmentation based on behavioural and engagement patterns
  • Predictive Modelling: modelling customer satisfaction and transaction activity

Developed for ISSS608 Visual Analytics