It is annoying when the user’s journey through the web store comes to an abrupt end with an error. Everyone is familiar with the frustration of online shopping: unclear purchase processes, non-transparent returns processes, pages that don’t finish loading, or simply errors in the applications themselves, such as when the data upload doesn’t work or the next page in the web store simply isn’t accessible.
The term user experience (UX) describes the experience that users have when interacting with a website or with specific content. Usefulness, aesthetics and stability determine the experience.
Usability, in contrast, is objective, the term describes the usability or rather the ease of use of the website or a piece of content during use.
In comparison to usability, user experience describes not only the experience during use itself, but also the expectations before and the perceptions after use. User Experience extends Usability by aesthetic and emotional factors. We have already analyzed some onliners honestly and relentlessly in this respect in the blog.
This article is about the reproducible errors in the application or the web store itself. If a print store operates more than one B2B store and one or more B2C stores by itself or on behalf of others, the company becomes a software hotbed.
Software developers try to ensure with regression tests that modifications in already tested parts of the software do not cause new errors (“regressions”). Modifications occur as a result of maintaining, modifying and correcting software. One builds up a scenario, documents the process step by step and repeats the test for each new software version. The more complex the software, the more extensive the test becomes, and at some point one working day is no longer sufficient. Developers are looking for a remedy in the form of automated tests.
“The term test automation covers all automated test activities and procedures used in the field of software development and the associated quality assurance. The developed software is thus tested as part of test automation using individually adapted testing tools. Such tools can either be programmed specifically for the software to be tested, purchased on the market as test automation tools, or in some cases used free of charge as open source variants,” writes the owner and founder of Critical QA, Florian Hemmann, on his homepage.
Whitepaper #4 “Test Automation” from T-Systems Multimedia Solutions provides an example. The authors assumed the following situation: A large online retailer operates several online stores, each with a different language. When new functions are implemented for one store, the adaptation for all other stores takes place automatically. Previously, the manual testing effort for a development feature was sometimes as time-consuming as the development itself, as the tests had to be carried out in the test, development and production environments. In addition, the recurring manual testing in different languages proved to be error-prone.
Through test automation with the test automation framework Testerra, automated tests can be executed in parallel for several stores overnight. It is also possible to use a test case in the same automated way for all other stores. In addition, a slimmed-down test set is available for hotfixes. Through regression tests and hotfixes it is possible to ensure continuous functionality of all stores. Furthermore, test data can be played out on all environments (development & test).
There are of course a lot of tools for desktop, web and mobile testing. Good results can also be achieved, for example, with Ranorex, a product originally from Austria. Ranorex Studio integrates with Jenkins, Jira, TestRail and other leading DevOps tools to automate bug tracking. The company says it supports more than 14,000 users worldwide in desktop, mobile and web application deployments. The test is prepared by recording actions with the Ranorex recorder. While it is necessary to redo the recording if anything fundamentally changes in the GUI, from personal experience the time saved is enormous. Not everything can be automated, certainly not if human factors are to play a role.
Testbirds takes a new approach and relies on crowd testing. Crowd testing is all about having digital products tested by real users under real conditions. By its own account, the crowd still includes 600,000 testers from 193 countries with more than 1,000,000 endpoints to test.
“Test automation overcomes a variety of challenges – but one myth that persists is the belief that software quality can be assured through test automation alone. Today, you have to include feedback from real users and consider the human factor when developing digital products. Involving the target group at an early stage of product development, for example through crowd testing, is crucial to ensuring that the customer experience is no longer a guessing game, but becomes a knowledge-driven strategy,” emphasizes Georg Hansbauer, CEO at Testbirds.
The cost of testing is great, but ultimately it serves to ensure the success of the business. Crowd testing is a new approach. Georg Hansbauer will be talking about it next week at the Online Print Symposium in Munich. I’m very excited about his talk, as he promises nothing less than to be able to take digital products to the next level, with real user insights from worldwide testers from his crowd. Good luck!