About the Client

MTN eShop is the digital commerce arm of MTN Group, a telecom giant operating across Africa and the Middle East. The platform enables consumers and businesses to seamlessly purchase data bundles, routers, security cameras, bulk SMS services, and licensed software. With millions of active users and high transaction volumes, MTN eShop demands uninterrupted service, secure data flows, and frictionless user experiences.

The Opportunity

MTN eShop was facing a silent yet critical issue that directly impacted user experience and trust. During a regression cycle, we uncovered a recurring flaw in the product review field: submitting special characters like <, >, &, or emojis caused the entire page to crash – without any warning or feedback.

What made this issue critical:

• User input was lost completely – no validation messages or error handling
  
• The crash bypassed manual testing and only surfaced with edge-case inputs
  
• The root cause was unsafe use of innerHTML, exposing the system to XSS risks
  
• Review abandonment increased, weakening a key element of product credibility
  
• Platform reliability and security were at stake, especially during high-traffic use
  

This hidden failure was quietly harming CX and could have led to bigger security exploits if left unresolved.

The Approach

Solving the issue wasn’t just about fixing a line of code – it was about reinforcing trust in a user-critical touchpoint. We needed to ensure the review system could handle real-world input patterns, edge cases, and potential security threats, all while maintaining platform stability. Our automation-first approach allowed us to act swiftly, diagnose deeply, and prevent future regressions – without slowing down MTN’s release cycles.

Here’s how we tackled it:

• Automated Edge-Case Simulation
  Enhanced our Cypress suite to inject special inputs – HTML tags, emojis, SQL characters, and multilingual content – directly into the review field.
  
• Root Cause Diagnosis with cy.intercept() + DevTools
  Pinpointed the use of unsafe DOM manipulation via innerHTML, which allowed unescaped user input to break the rendering logic.
  
• Scalable Test Data Using Faker.js
  Generated randomized and complex user inputs, including long strings and invalid characters, enabling repeatable and high-coverage QA cycles.
  
• Secure Coding + User Feedback Layer
  Worked closely with MTN’s dev team to:
  
  • Replace innerHTML with textContent or safe rendering methods
    
  • Sanitize inputs at both frontend and backend layers
    
  • Implement real-time validation to prevent crashes and guide user input

The Impact

What began as a subtle UI flaw in MTN eShop’s product review system had the potential to compromise customer trust, data integrity, and platform security. By addressing this vulnerability with automation-first QA and precise input validation, Ekfrazo helped transform a high-risk edge-case into a stable, scalable user experience.

Review submission failures – previously affecting nearly 30% of users – were brought down to zero. Page crashes caused by special character inputs were fully eliminated, and critical user input is now preserved with real-time validation. Manual regression testing, once consuming over 2 hours per release cycle, has been replaced with robust automated test coverage.

This solution delivered more than just a fix – it created measurable business value. We strengthened a CX-critical module without rewriting the platform, reduced QA overhead through negative test automation, and eliminated silent friction points affecting telecom eCommerce conversion. With QA now aligned to real-world behavior, MTN eShop gains improved platform resilience, accelerated release velocity, and renewed customer confidence.