Shulou(Shulou.com)06/03 Report--

A successful software must add value to the enterprise business by simplifying the process, so enterprises must think about the following questions when or even after the software selection: can the software project bring the desired improvement? Will users get a better experience? Will the productivity be increased as expected? Are users satisfied with the changes and consistent with previous expectations?

In order to ensure that the software can achieve business goals in terms of workflow, user satisfaction, design strength, and so on, we need an artifact-usability testing. So, how do you choose usability testing? When and which usability tests should be used? This article will introduce you to six kinds of usability testing in detail.

A / B test

The A / B test applies to a website or landing page, while testing two separate designs (An and B) over a period of time and collecting data about their performance. If A's test result is better, then we choose to use A, and then test the other parts in turn. Such usability testing usually requires the assistance of third-party tools, and there are many third-party solutions.

How should a good A / B test be performed? In order to better compare the specificity between each design, it is best to change only one element at a time, clearly defining test goals, user scenarios, problems that users may encounter, and so on.

List of changed elements:

Define specific issues, such as the user's failure to complete the process of submitting the form.

Define specific goals, such as increasing user form submissions.

Define tests, such as design A can be the location of the submit button, while design B can be a different location.

Each design is tested for a period of time through a specific form of measurement, such as a three-week test on 500 users.

The declaration of the test results may be as follows:

Over a period of three weeks, we conducted two tests with an average of 500 user interactions, with an average number of users per design. The test results show that the submission rate of the blue button is 46%, while the orange button is as high as 76%. To sum up, the design of orange button is more suitable.

The Ahand B test is efficient and clear, and it takes some time to see the results. In addition to univariate, multivariable tests can also be in a format similar to the Ahand B test, but the test results may not be as accurate as the Ahand B test.

Prototype testing

Prototype testing is early testing, which usually tests the complete user workflow before development, and the guidelines to follow before prototype testing begins include:

Define the budget and goals for the test

Rely on early testing to identify specific areas that may need improvement

Select a prototype tool, such as an Axure solution

Select a prototype measurement tool to collect user analysis, and the team that manages the test must be familiar with the tool and learn how to measure tests, such as Loop11.

Formative usability testing

Formative usability testing is another form of early testing, focusing more on quality, and the product needs to pass the acceptance standard test before it is released to the target user. The following process is followed to form a usability test:

The product can be soft-started before release (v.0.5).

Perform defined usability tests with the beta test group.

Write specific test cases to guide users to achieve specific test goals.

Select a third-party tool, such as Optimizely, to help run tests and collect analysis.

Review and analyze and make business decisions for product design.

Modify the design and resolve usability issues before the product is officially released.

Run other tests to continuously improve the product.

Summative usability testing

The summary usability test is carried out in the later stage of development, and the goal is to determine whether the design is really in line with the goal of the product, the test object is a defined user group, and the result should be an accurate statistical measurement of usability. Make guiding decisions for the testing phase. Summary usability testing usually follows the same process as formative usability testing, but pays more attention to detail and user experience, which should be consistent with each new version of the product.

Eyetracking technology

Eyetracking technology is a very special type of test, which tracks the user's eye movement and gaze when using the product or browsing the website. It usually requires the help of third-party software tools, such as iMotions.com, to track user clicks and mouse movements. This type of testing is often performed in 10 or more defined user groups, analyzing the data and determining how to improve design and UX performance. However, the cost of this kind of testing is very high, which is an important reason for the low usage of such usability tests.

questionnaire

Questionnaires are not as numerically based and accurate as other forms of testing, but they provide basic feedback from user groups, and questionnaires are the most economical solution if you want to collect a large amount of information in a short period of time. and easy to use, even inexperienced small test teams can do it.

The questionnaire can also be executed by third-party tools. The biggest difference between the questionnaire and other forms of usability testing is that it has formulated a series of opinions, but this approach also has shortcomings. As long as any point of view reaches a certain number, it will affect the validity of the results, and the questionnaire, which is more inclined to the subjective intention of the questioner, is easy to get wrong conclusions.

In summary, usability testing is a very important thing for software development, and early testing can not only improve the success rate of the product, but also save time and money.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.