I am into Voice UI because it is the future of product design and it is challenging. It is extremely difficult to create a useful conversational experience because the front part of our brains are wired for visual processing. And we have not made very far in making users feel like they are interacting with a human - not a machine - in today's VUIs.
VUI does not need to have a visual interface - it can be completely auditorial or tactile. This can be frustrating because a human brain can easily process a map interface that shows your 6 nearby grocery stores, but it is a pain to translate the same amount of information into a conversational experience.
There have been a ton of resource poured into developing VUIs, but most products still fail to cultivate trust and ease of use between users and AI. This failure proves to be the most expensive mistake that any company can make. Why? Because VUIs might take longer to ship and therefore cost more money. Many of today's first VUI use cases actually deal with IoT, which involves processing huge amount of data. And the OKRs set by the organizations to evangelize the grand vision seem to be completely detached from understanding actual users at the other end.
How do designers help? The true value is that Product Designers can help Product Managers conduct continuous discovery and test assumptions before diving into engineering. The optimal PM - PD relation results in an efficient workflow where designer can create prototypes based on requests and comments from users, and evaluate feasibility with light-weight experiments both internally with developers and externally with customers.
In the case of cross-platform VUIs - let's be honest - the requirements are more than complex. And the value can be brought in by designer is big. By doing an end to end tracer bullets prototype (no code yet! just design) at this stage when all the stakeholders are present can help square away major UX issues if Product team assembles the right reference customers.
One example is that the first thing Voice Assistant should do right is to pick up user's name. Voice Assistant should pick up the right cultural context and use the appropriate accent to utter user's name, based on training data that comes from native speakers that speak the same language(s) as the user. This would be one of the first things a good Product Designer notices during testings of the state-of-art conversational products.
In the end, a good Product Designer is well-versed in testing methods and knows when to ask what types of questions to get the information needed. There are multi-variant testing, value testing with customers, and internal testings with engineers while getting them on board. Once Product team establishes a good cadence of continuous discovery workflow with both internal developers and external customers, they can capture and tackle a surprisingly large amount of issues pre-development! At the end of of day, companies can save a ton of money and successfully bring better VUI products to market with confidence.