Voice User Interface (VUI) is a speech recognition technology. It allows people to interact with devices and applications using their voice. Apple's Siri, Google's Assistant, Amazon's Alexa, and Microsoft's Cortana are prime examples of VUIs.

Let’s face it humans are social animals. We need interaction in some form, and voice being our primary mode of interaction makes VUIs one of the fastest-growing technological advancements in this age and time. The voice-first approach allows users to fulfil their day-to-day tasks and initiated automated services in a faster and more intuitive way.

Usability studies have shown that VUIs, generally have poor usability and are limited to simple queries that have very simple short answers.

As VUIs are rapidly growing in capabilities, it becomes more important than ever to test your VUIs and make sure your products are not only solving people’s needs but doing it in a way that keeps the products customer-centric.

While voice-based products are technical marvels. However, they bring their own set of problems when it comes to testing

• Tools: Tradition testing methodologies and tools are text and visual-based or are not equipped for voice.
• Scale: While inputs in a traditional program can be limited. Voice can have infinite variations and complexity (eg. Language, accent, pronunciations etc. )

Testing traditional applications is easier when compared to testing voice-based applications for voice-based application testing you need real voices, and real people, there can be multiple ways to reach the same point. “Alexa, Will it rain today in Mumbai?” “Alexa, Will it rain today?”

While the first statement is giving all the details needed to get a result, the second statement is vague, however, both these should be handled in a way where the user does not have a feel like the application isn’t able to understand their needs.

Approach to VUI testing