Optimizing Furniture Platform Stability & Performance with Manual and Automated Testing

The Client

Our client is a leading furniture aggregation platform, integrating and presenting thousands of products from multiple retailers in one place. By offering a seamless search and comparison experience, their platform enables users to explore a wide range of furniture options from different brands and retailers without having to browse multiple websites. Given the nature of their business, ensuring high platform reliability, fast performance, and accurate data ingestion is critical to maintaining a smooth user experience and strengthening trust with both customers and retailers.

 

The Challenge

As the online furniture retail market grows, aggregation platforms face increasing challenges in handling vast amounts of data, managing complex integrations, and ensuring platform stability under high-traffic conditions.

Before implementing a structured testing process, the platform experienced:

• Intermittent failures during traffic spikes, affecting user experience.
• Bugs related to retailer data ingestion, leading to incorrect product listings.
• Quality control issues due to the fast-paced continuous delivery process, making it difficult to identify issues before new releases.

To address these issues, our client recognized the need for a dedicated QA strategy that would introduce both manual and automated testing, ensuring faster bug detection, improved platform resilience, and an overall more stable system.

This led to the integration of a dedicated QA expert who worked alongside the development team, embedding quality assurance into the development workflow while ensuring minimal disruption to deployment speed.

 

The Solution

A skilled QA professional was introduced to design and implement a hybrid approach, blending manual testing for critical workflows and exploratory analysis, and automated testing for continuous quality assurance and performance optimization.

1. Manual Testing for Immediate Quality Assurance

Manual testing played a crucial role in identifying real-world usability issues and mitigating risks before automated testing was fully implemented. Key initiatives included:

• Exploratory testing: A hands-on approach was used to uncover UI inconsistencies, broken links, and logic flaws that automated scripts might miss.
• Usability validation: The QA specialist reviewed site navigation and checkout flows, ensuring a seamless and intuitive user experience.
• Retailer feed validation: Since the platform aggregates thousands of products from various vendors, manual testing helped verify that product descriptions, prices, and availability were accurately reflected on the platform after catalog updates (ETL).

2. Automated Testing for Speed and Scalability

To ensure efficiency, reliability, and faster deployments, automated test cases were introduced across key areas:

Automation of important flows: It was implemented automated scripts to test important and critical flows, simulating a real user journey, such as product search, filtering, image search, customized website components, etc.

• In global, there are around 900 points of verification.
• The full end-to-end battery of tests execution, in parallel, takes around 45 minutes to execute.
• The full battery of the automated end to end tests is executed twice a day.
• For each pull request made into the frontend repository, a group of smoke tests is executed.

Automation of internal API tests:  For each internal API used by the company systems, it was implemented automated integration tests.

• Full batteries of integration tests were implemented along 5 distinct APIs.
• For each pull request made into the API repository, the full battery of the automated integration tests on the scope of that repository is executed.

Performance testing: It was configured automated tests to simulate traffic spikes and ensure stable platform performance.

• These performance tests helped improving the platform stability and performance. The work done in the beginning was interactive, where adjustments in the infrastructure were made upon the performance tests results and then the tests were executed again, new adjustments were made, until an optimal result was reached.
• The performance tests also helped identifying bottlenecks. A real user journey was executed for many hours, helping to identify bottlenecks and parts of the code where the memory handling needed to be improved.

Continuous monitoring: Established a continuous monitoring system to detect and resolve issues in real-time. For this, the QA built GitHub workflow pipelines to automatically execute the tests periodically or upon certain events.

3. A Collaborative, Agile Testing Process

Beyond implementing testing mechanisms, the QA professional became an integral part of the development process, ensuring quality assurance aligned with business priorities. The approach included:

• The QA worked closely with developers to identify priority areas for testing and suggest improvements.
• Provided regular feedback on product quality, helping the team deliver more robust and reliable versions.
• Working in a continuous deployment system, speed is key, so the QA professional adapted the QA processes to diminish complexity and bureaucracy but still maintaining the high degree of process quality.

To ensure the project’s success, 99x Europe embraced a structured and collaborative approach with the client:

• Daily standups: To keep all teams aligned on progress and priorities.
• QA process: To guarantee that each feature met the client’s high standards through rigorous testing and validation.
• Team spirit: To foster an environment where every member could count on support from their peers, ensuring that challenges were overcome together.

 

The Results

The introduction of manual and automated testing had a significant impact for the platform, for users and for retailers.

For the Platform:

• Bugs reduction: There was a decrease in the number of bugs reported by users after implementing the testing strategy. Overtime, the confidence when deploying code changes into the production system increased. Since 2022 the system uptime is in average 99,99%, exceeding the original 99.90% target.
• Improved performance:  The platform can handle the expected traffic without compromising speed or stability.
• Consistent quality: Updates are launched with greater confidence due to thorough pre-release verification.

For Users:

• Enhanced experience: Users now experience a smoother and more reliable experience, with fewer interruptions during navigation.
• Increased trust: Due to the tests and fixes done in the retailer’s data ingestion process (ETL), the accuracy of displayed data (prices, availability, descriptions) was significantly improved, boosting user confidence.

For Retailers:

• Stable integrations: Retailer feeds are now integrated with greater precision, reducing the risk of errors and increasing the synchronism between their data and the information displayed on the aggregated platform.

 

Technologies

Programming & Infrastructure: JavaScript, TypeScript, C#, Azure, SQL Server, Cosmos, Kubernetes, Service Bus, GitHub.
Testing & Automation: Cypress, Cucumber, SpecFlow, Jest with Supertest, Grafana K6.

 

Investing in manual and automated testing has transformed our client’s platform into a highly stable, scalable, and user-friendly solution.

 

If your business is also looking to elevate product quality, optimize testing processes, and enhance platform reliability, get in touch with us. Let’s build something bullet-proof together.