<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=278116885016877&amp;ev=PageView&amp;noscript=1">

Jan 7, 2024 | 4 Minute Read

How Quality Engineering Solves Challenges In Building A Feedback System


Table of Contents


Understanding your users is crucial for organizational success. A robust feedback system gathers data ( through reviews, surveys, social media, etc.) to unlock valuable insights. This identifies key areas for product improvement and informs data-driven decisions that optimize future offerings. 

The redesign of Airbnb, motivated by customer feedback, is an example of how well this principle works.

On May 3, 2023, Brian Chesky, the CEO of Airbnb, announced a significant platform redesign featuring over 50 improvements to enhance its customer experience. This initiative was driven by extensive user feedback gathered from diverse sources, such as:

  • Guest reviews
  • Host input
  • Surveys
  • Social media
  • Direct channels

Acting upon it enabled Airbnb to 

  • Understand user needs and trends
  • Improve functionality, usability, and platform performance
  • Prevent negative experiences and reduce operational costs
  • Increase host satisfaction and reduce guest complaints
  • Build trust and strengthen the brand's reputation

Airbnb's success hinges on its strong feedback system. To replicate these results, organizations need to solve challenges in setting up a strong feedback system. 

Quality Engineering (QE) offers valuable assistance in overcoming these hurdles.

Challenges Setting Up A Strong Feedback System

Data Processing

Data processing in feedback systems is about turning raw data into valuable insights by cleaning, classifying, and analyzing it. It's sorting and refining messy input to reveal hidden patterns and trends, a key ingredient for continuous improvement.


Lack Of Qualitative Data (Non-Numerical Information & Concepts)

Usability testing and user research can help build context by sharing insights on user motivations, behaviors, and pain points. 

Poor Data Quality

The QE team automates data quality tests integrated into the CI/CD pipeline. During product implementation, the team identifies anomalies in historical data through close collaboration and implements relevant scenarios into the test suite, ensuring comprehensive testing.

According to Gartner, each year, poor data quality causes organizations to lose an average of $12.9 million. Besides the direct impact on revenue, it complicates data systems and results in bad decisions over time.


Simple interfaces, clear instructions, and quick processes encourage more people to share their thoughts, providing valuable insights for improvements and better outcomes. The usability of a feedback system has to be top-notch.


Creating a Good Experience

Data about user actions and emotions can be used to create an empathy map. These maps guide design, ensuring each iteration resonates with real users, ultimately leading to delightful experiences. 

Feedback System Design

A good A/B test helps determine the preferred interface. Testers can assess user preferences by comparing variations and making data-driven decisions on visually appealing and impactful changes.

Netflix maximizes user engagement and revenue growth in the competitive streaming industry through a data-driven strategy. At the core of this approach is the extensive use of A/B testing. 

Netflix optimizes content recommendations and personalizes the user experience by continuously iterating and testing features and algorithms. 

This emphasis on A/B testing allows Netflix to refine its offerings based on real-time insights into customer preferences.


Scalability in a customer feedback system refers to its ability to efficiently handle growing volumes of feedback data, users, and interactions. It involves seamlessly adapting and expanding the system to accommodate increased demand without compromising performance or the user experience.


Feedback Volume Surge

Conducting load testing by mimicking high user loads reveals performance bottlenecks within a system. This process aids in identifying areas for improvement, allowing optimization of server capacity to enhance overall performance and the user experience.

Security Concerns

A growing user base attracts more security risks. To protect the system, perform security audits, conduct penetration testing, and enforce strict access controls, ensuring robust defenses against potential vulnerabilities.

User Interface Responsiveness

Data surge poses the risk of slowing down the user interface. Frontend performance testing enhances code efficiency, reduces rendering delays, and ensures a responsive user experience.

Meta's AI training experienced bottlenecks due to the huge amount of data that slowed down their CPUs' data processing, which lagged behind the lightning-fast GPUs used for AI-model training. 

This and limited power budgets restricted the number of models they could run.

Meta implemented a comprehensive strategy to address these challenges, including building a new data ingestion infrastructure and last-mile transformation pipelines. A key component is the Disaggregated Data PreProcessing Tier (DPP), responsible for fetching, decoding, and transforming data for AI training. The DPP allows scalable data ingestion and independent scaling.

The optimizations led to a 35–45% improvement in the power budget required for data ingestion, enabling support for a growing number of AI models within power constraints.

Organizational Challenges

Organizational challenges in setting up a feedback system include securing stakeholder buy-in, defining clear objectives aligned with business goals, addressing potential resistance to change, and establishing a culture that values data-driven decision-making.


Stakeholder Buy-In

Generating a comprehensive report showcasing tangible benefits can positively influence stakeholder buy-in. Utilizing QE metrics to define SMART (Specific, Measurable, Achievable, Relevant, and Time-bound) goals for the feedback system is also convincing.

Fostering A Data-Driven Culture

QE values data as a learning tool for guiding informed decisions and ensuring inclusive products. With clear goals, shared celebrations, and stakeholder involvement, QE sets the stage for a data-driven culture. 

An example of a tangible benefit:

When it comes to the Cost Per Ticket (CPT), IT companies rank among the most expensive. The CPT value typically ranges from $25 to $35, but it can reach up to $100 depending on operational factors.

In a scenario with 500 monthly support tickets at $35 each, a 20% reduction through efficient QE results in 400 tickets.

Monthly cost savings amount to $3,500 ($17,500 - $14,000).

Data Privacy, Security, & Compliance

Establishing a feedback system involves navigating data privacy, security, and compliance challenges. Safeguarding user information, implementing robust security measures, and ensuring compliance with relevant regulations are critical aspects of creating a trustworthy and reliable feedback platform.


Identifying Vulnerabilities 

Security testing identifies vulnerabilities and establishes robust encryption protocols for secure customer data handling. Regular audits validate compliance with data protection regulations, ensuring the integrity of the feedback system.

Protecting Users' Privacy

Collecting only essential information, implementing secure storage with encryption, employing access controls, being transparent with users, complying with privacy regulations, and conducting regular audits contribute to privacy.

A 2023 survey by Deloitte claims that 50% of customers think that the benefits they get from online services are not worth worrying about data privacy. This highlights the importance of organizations showcasing their commitment to robust data privacy measures.

Businesses must clearly communicate their data practices, secure handling, regular audit details, and other security measures through various mediums. 

How Canva Made Their Customers and Team Happy with QE Approach 

Canva strategically enhanced its quality model despite resource constraints and tight deadlines. 

Their new Quality Assistance model embraces a proactive "shift-left" approach that involves addressing potential challenges at the early stages of development to prevent them from escalating. 

They leveraged data-driven decisions to optimize their quality process, which included

  • Prioritizing testing for the most-used features, which minimized critical bug encounters and boosted overall satisfaction.
  • Focusing on code coverage improved test efficiency, reduced reworks, and increased release confidence.
  • Balancing technical debt with new features, preventing future issues, and intuitive user experiences.
  • Understanding user behavior helped test high-traffic areas, reducing user friction points.

As a result, their team achieved better test coverage, reduced code rework, and increased confidence in feature releases, resulting in zero incidents in the past three months.

If you're eager to delve deeper into how QE can elevate your systems and processes, why not schedule a call with one of our QE experts?

About the Author
Shweta Sharma, Director of Quality Engineering Services
About the Author

Shweta Sharma, Director of Quality Engineering Services

When Shweta isn't at work, she's either on a family road trip across the country or she's dancing with her kids—it's a great combination.

Vishnu S Kumar, Marketing, Associate

Vishnu S Kumar, Marketing Associate

A humble and passionate soul who loves to explore the world on his bike—that’s Vishnu. In his free time, you’ll often find him at cafes, sipping coffee and watching different genres of shows.

Back to Top