Why Usability Should Be Top Priority for Critical Systems

It’s a truth universally acknowledged that things need to be user-friendly in order to succeed. Whether it’s an app on your smartphone or a medical device keeping patients alive, a poor user experience is a sure-fire way to fail. But how do we ensure this isn’t the case? Three words answer this question: test, test, test. 

The Importance of Being User-Friendly 

Any system we use in our daily lives should aim to be user-friendly. There’s nothing worse than downloading an app onto your phone and finding the interface to be cluttered, complex and difficult to use. Most of the time, a situation like that will lead just as quickly to uninstallation! 

But the usability of critical systems means much more than a slightly annoying app. It could mean the difference between life and death, or between the success and failure of a business. Depending on the system, a lack of usability testing could result in spiralling project costs, copious design flaws and potential non-compliance with industry regulations, meaning the final product may never come to market at all. Whatever the costs of usability testing now, it’s clear to see that it’ll be worth it in the long run! 

To use a slightly less fatal but certainly no less important example than our pilot scenario, consider a retail bank that wants to introduce a new app for their customers, allowing them to manage many of their money matters on their phones. The services available on this app would include the usual activities of transferring money between accounts but may also cover mortgage applications and loan eligibility checks. 

However, no usability testing was performed on the app at any stage of the development process. This meant there was no complete UxDesign process, with the software developers simply assuming that users would be able to navigate the app and complete their tasks without a hitch. 

Because of this lack of consideration for UxD, the layout of the app’s home screen is massively overburdened with various portals to the different services available, causing confusion amongst users who simply want to access the services they need at any particular time. When users finally figure out how to locate the service they need, the input field labels for data required to check loan eligibility are not clear, meaning many users are either put off at the first hurdle or end up inputting incorrect data and hence producing inaccurate results. 

The complexities of this process, and the less-than-satisfactory outcome produced if and when the customer makes it to the end of their journey, will more than likely push said customer away from this bank and move to a competitor whose understanding of customer needs is more mature. It’s therefore clear to see just how business can be affected by not paying appropriate attention to UxD, even before the development process for a new product or service begins. 

Certainty Through Usability

We’ve determined just how important usability testing is – but how does it work, exactly? 

Usability testing does not mean sitting a focus group of people down in a room and discussing their opinions on a product or service. Usability testing is actually putting the product or service in the hands of the potential user, giving them the opportunity to discover and explore it. 

These tests can be moderated or unmoderated, which both have their own pros and cons. A moderated test may allow for greater insights into the in-the-moment reactions of users to the product/service under test, although could increase testing costs. By contrast, an unmoderated test could be useful to keep the project within budget, although the lack of real-time support for users could compromise their understanding of the product/service, especially if the test subject is a prototype or a series of wireframe diagrams.

AI-ming for Great Usability

This year’s World Usability Day places a special focus on the possibilities opened up to UxD by artificial intelligence (AI). It’s estimated that up to 95% of customer interactions will be AI-powered by 2025. And with the rise of products and services which rely on AI to function comes new user experience considerations. For instance, how do we test a product/service which does not necessarily rely on human input to perform its functions, but rather runs its own processes and generates its own results? 

Two key requirements of AI-based products/services are reliability and trust. The user must be able to trust the product or service they are working with to overcome one of the first hurdles in ensuring product/service success: uptake. A critical system that cannot be trusted or isn’t deemed reliable enough is undoubtedly doomed to failure before it’s even got off the ground. 

But it’s not just AI-powered products and services that usability testing has to contend with. AI may well be the future of usability testing, too, replacing human testers altogether. A recent study assessing the ‘tappability’ of 3,500 apps, covering 20,000 in-app elements and ranking them based on how easy they are for users to interact with. Samples results were fed into a neural network which then produced a binary result of the ‘tappability’ of each app element (where ‘1’ was a tappable element and ‘0’ was not-tappable). A control group of 290 human volunteers was used to carry out the same assessment of app ‘tappability’, albeit covering a smaller number of app elements. The results of this study demonstrated clearly that the difference in test outcomes between the AI neural network and the human testers was marginal, the mean precision being 90.2% for both groups. The inconsistency of labelling between groups was also equal, at 40% of the elements assessed. 

AI could be the future of usability testing, reducing costs and achieving similar, if not more accurate results than human participants.

To discover more about how Critical Software puts usability at the heart of its software development, check out our free UxD pocket guide. Happy Usability Day!