Customer Support Transformation: Improving Customer Experience Through CRM Implementation

Why I worked on this project

One thing I've noticed while learning data analysis is that many portfolio projects stop at reporting.

You analyse the data.

Build a dashboard.

Present the findings.

But in reality, businesses don't invest in dashboards.

They invest in solving problems.

For this project, I wanted to go one step further and combine Business Analysis with data analytics to simulate a real business transformation.

The business problem

Imagine a customer support team working without a central CRM system.

Customer information is scattered.

Agents duplicate work.

Response times become inconsistent.

Managers have limited visibility into performance.

Customer satisfaction begins to fall.

The problem isn't simply that reports are missing.

The problem is that the organisation lacks the information needed to improve its processes.

My approach

Instead of jumping straight into the data, I approached the project as a Business Analyst would.

I started by understanding the business problem before looking at the numbers.

The project included:

• analysing the current support process (AS-IS)

• designing an improved future process (TO-BE)

• identifying stakeholders

• defining business requirements

• prioritising requirements using the MoSCoW method

• preparing user stories and acceptance criteria

• documenting business rules and risks

  
  

Only after establishing the business context did I move on to the analytical work.

Data analysis

Using SQL, I analysed key operational metrics such as:

• ticket volumes

• SLA performance

• average resolution times

• customer satisfaction

• agent performance

• monthly trends

  
  

Python was then used to explore patterns, identify relationships between variables, and better understand operational performance.

Finally, I developed an executive Power BI dashboard designed to support management rather than simply display numbers.

What the analysis revealed

Several recurring issues became clear.

High-priority tickets often took longer than expected to resolve.

Some agents carried significantly heavier workloads than others.

Customer satisfaction tended to decline as resolution times increased.

Certain ticket categories generated a disproportionate share of support demand.

Perhaps most importantly, the lack of centralised reporting made it difficult for management to identify these patterns quickly.

Business recommendations

Rather than ending the project with a dashboard, I focused on potential business improvements.

The recommendations included:

• implementing a CRM platform

• introducing automated ticket routing

• improving SLA monitoring

• standardising customer communication

• balancing workloads across agents

• introducing executive KPI reporting

  
  

The goal wasn't simply to describe what had happened.

It was to suggest how the organisation could improve.

What I learned

This project reinforced something I've written about before on this blog.

Data analysis becomes far more valuable when it is connected to business context.

The dashboard is only one part of the story.

Understanding the problem, gathering requirements, analysing processes, and translating insights into recommendations are what turn analysis into business value.

Tools used

• Business Analysis

• SQL

• Python (Pandas, Matplotlib)

• Power BI

• Microsoft Word

• BPMN

  
  

GitHub

The full project, including business analysis documentation, SQL scripts, Python notebook, Power BI dashboard and supporting files, is available on GitHub.