A Hybrid Usability Approach for Rating Evaluation of M-Commerce Applications

In the bustling world of modern shopping, the difference between a successful app and a forgotten one often comes down to a single, fleeting moment: how easy it is to use. This field of study, known as usability, examines the friction between a person and the technology they hold in their hand. It asks simple but profound questions: Can a user learn the interface quickly? Does the system behave consistently? Is it satisfying to complete a task? For businesses selling goods through mobile devices, these questions are not just about comfort; they are about survival. When an application is difficult to navigate, users leave, and the digital storefront loses its value. Understanding which specific design elements drive a high rating is crucial for developers who want to build tools that people actually want to keep.

A team of researchers in Pakistan set out to decode this relationship by looking at five popular mobile shopping applications used in their region: Daraz, Shophive, Home Shopping, Symbios, and Yayvo. They began by gathering a wide range of ideas about what makes software easy to use, drawing from established theories that break down usability into eight distinct categories. These categories included how quickly a user can learn the app, how consistent the design feels, how well the system handles human error, and how efficient the overall experience is. To refine this list, the researchers consulted a panel of eight experts in the field. Through a careful process of discussion and revision, they narrowed down a long list of potential questions to a focused set of sixty items, ensuring that every question measured something meaningful about the user's experience.

With their refined questions in hand, the researchers turned to the people who matter most: the shoppers. They recruited two hundred individuals who regularly used these mobile applications to buy products. Each participant was asked to perform specific tasks on the apps and then answer the survey questions, rating their experience on a scale from strong disagreement to strong agreement. After removing incomplete or inconsistent responses, the team was left with one hundred and sixty-eight detailed accounts of real-world usage. This data became the foundation for their investigation, allowing them to see which factors actually correlated with the ratings users gave to the applications.

The researchers first looked at each of the eight usability factors on its own to see how much it influenced the final rating. They found that efficiency—how fast and smoothly a user could complete a task—was the strongest individual predictor of a high rating. Conversely, they found that general human factors, a broad category covering the physical and psychological fit between user and machine, had the weakest influence on the rating when considered alone. This initial step revealed that not all aspects of usability are created equal; some matter far more than others in the eyes of the consumer.

To get a clearer picture, the team combined these factors into a single predictive model, using a statistical method that builds a formula by adding variables one by one until the best fit is found. They discovered that a model using seven of the original eight factors could explain ninety-one percent of the variation in user ratings. The one factor they had to leave out was the "human factor," suggesting that while it is important, it does not add enough unique value to the prediction when the other seven factors are already accounted for. The resulting model showed that learnability, efficiency, and effectiveness were the heavy hitters, carrying the most weight in determining whether an app would be highly rated.

The team then compared their new, streamlined model against a simpler approach that treated all eight factors as equally important, essentially averaging them out to guess the rating. The new model, which carefully weighed the importance of each factor, proved to be vastly superior. While the simple average explained only a tiny fraction of the rating differences, the new model captured the vast majority of the story. To ensure this result was not a fluke, the researchers tested their model repeatedly by splitting the data into different groups, a process that confirmed the findings were stable and reliable across different sets of users.

The study concludes that for mobile commerce applications, success is not about checking every possible box of usability. Instead, it is about mastering a specific set of core attributes. By focusing on how quickly users can learn the system, how efficiently they can complete their shopping, and how effectively the app helps them achieve their goals, developers can significantly improve the ratings their applications receive. The research suggests that a targeted approach, honing in on these critical seven factors, offers a much clearer path to creating a successful mobile shopping experience than trying to optimize every single aspect of the design equally.