Blog

Things to consider when designing a voice UI

Feb 19, 2019 8:00:00 PM

Voice UI

Increasingly more voice-powered gadgets, like the Google Home, Apple HomePod, and Amazon Echo are raging the market. It would not be a revolutionary explanation to state that refined AI will be in each home and on each phone soon. Regardless of whether we're discussing VUIs for smart home speakers or mobile applications voice communications are getting progressively common in the present technological era, particularly since screen fatigue is a worry.

In any case, with the multiplication of AI comes an influx of doubt, as we combine new and diverse devices. When the users can't confide in an AI's Voice User Interface, they will be probably not going to associate with it or purchase anything from it.

So, building a Voice User Interface (VUI) that is smooth, natural, and consistent is of foremost significance to AI organizations that intend on staying around.

Let’s look at some of the important strategies of designing a voice UI.

Concentrate on the user

Concentrate on understanding the user’s purpose (a reason behind interaction). At the point when users communicate with a device, they have a specific issue they need to the solution for, and the objective of the designer is to comprehend what this issue is.

Designing a voice-first user experience and creating techniques that are natural and satisfy users require a somewhat unconventional outlook, yet it generally starts by giving an attention on the user and the issue they're endeavoring to unravel.

Think about the condition the user will be in, what context they will have advancing on, and what it is they're endeavoring to do? Ensure you consider the ideal approaches to keep things moving as you make your experience.

When you structure, prototype, and distribute so as to test the voice experience, ensure you have concentrated on solving the correct issue. Your design and prototyping time is sufficient to make an engaging experience.

Designing for context and accent

One reason numerous voice-empowered devices - just as chatbots - as of now battle with conversational UI is since they need context. Context implies monitoring what's happening around the conversation, just as things that have occurred previously.

Regularly the wording utilized amid the discussions may have distinctive implications for various contexts given. The framework should be instructed for these contexts circumstances and the dialog exchange ought to be driven by contexts.

In terms of the accents, there are various accents of English in the world. So, at this point envision the scale when you are building it for the world. The app ought to have the potential to select these accents and furthermore learn from the examples.

It implies VUI designers ought to dependably consider about potential hearing impairments, speech obstacles, and whatever other things that could impact correspondence, for example, cognitive disorders. Indeed, even accent, a tone of voice, or dialect can influence how the gadget comprehends them.

It's majorly a learning process, and it additionally customizes the app for a person. It takes certain time, however, with the available toolboxes it's certainly achievable.

This expects you to be smart about how and where to utilize voice and design it such that everyone can utilize it, regardless of how they hear and how they sound.

Dialog management

Dialog management is the manner by which the framework oversees what has just happened in the conversation and what necessities to occur subsequently.

To make your VUI as adaptable as could be reasonable a concept termed dialog management gets critical.  It is vital as the users need to give data in various ways; some will offer data bit by bit, and somebody will offer all the data up front. To deal with this, the application has to recognize what bits of data are required, these bits are some of the time alluded to as slots - and in what order to request them.

Make it conversational

It is important to ensure that a voice UI comprehends natural speech – for example, it can acknowledge a wide scope of various inputs. We talk in a different way as compared to how we type, in questions or full sentences, as opposed to a progression of keywords.

Envision your Saturday night, you take out your smartphone and type in it – “a club nearby." Predictably, a list of results will show up on your screen.

But when we communicate with a voice-powered device, you'd be likely to express your request in this way - "Alexa, can you suggest a good nearby club for a party?"

Thus, machines must have the capacity to comprehend and react to the diverse commands to be fruitful.

Create visual mockups

At the time of designing a mobile application, mocks/wireframes are obviously an essential part of your initial design process for a VUI application — they'll run simultaneously with sample dialogs to help picture the user experience. Your mockups/wireframe are your storyboards: it's critical to collecting them.

In case the VUI staff is separated from the visual staff, ensure you meet up for this piece. For the user, it's one experience, so, visual designers and VUI designers must cooperate firmly in early stages.

Make a use case metrics

What is the tertiary, secondary, and primary use cases for the VUI? Does the gadget have one primary use case? Or on the contrary, does it have a diverse blend of use cases?

It is critical to making a use case grid that will enable you to distinguish why users are collaborating with the gadget. What is their essential method of cooperation? What is optional? What is a decent to-have collaboration mode and what is fundamental?

You can make a use case metrics for every mode of communication. At the point when connected to voice interaction, the framework will enable you to see how your users as of now use or need to utilize voice to interact with the product, incorporating where they would utilize the voice assistants.

User navigation

Voice interfaces are the same and users may end up lost amid their journey, giving them an exit plan gets something a troublesome undertaking with the VUIs. Help them get support with an appropriate recommendation on the actions to be performed, and this will likewise keep the odds of the user making mistakes.

Limit the amount of information

With verbal content, you require to keep all information and sentences concise so that the user doesn't become disoriented or neglect items on a list. Amazon recommends limiting the options to three for an interaction.

In case you have a more extensive list, you must make the group of options and begin by offering the users with the most eminent ones. Inform them that they can request more choices by asking if they would prefer to have more options.

But user negligence is only half the problem here. The other side of the coin is that the disappointment that will heat up within them when they get caught on a long, recoiling confusing route when all they need to do is reach a simple purpose.

Keeping the persona predictable outside the VUI

In case the graphical components don't agreeably supplement those of the voice, making an effective VUI will end up being a non-beneficial attempt.

A huge consideration must be given in reflecting qualities of your conversational operator’s voice to its content and other identical visual representations, such as gifs or symbols. Is your operator’s persona to be friend-like and easygoing? So, any notifications or responses it shows in content, either spoken or not, must line up with that. A remiss methodology on grammatical conventions, for example, punctuation, syntax, and even slang words would be an easy decision all things considered.

On the contrary, how is your operator’s responsiveness depicted in an absolutely graphical sense? For example, blending an excessively cartoonish icon with a highly-functional smart assistant could infer to users that it's a gimcrack instead of a utility.

Personalization is principal

As of now in voice-enabled gadgets, there's a liberal inclination towards personalization. It is all vital for making an efficient, quick experience. An incredible method to do this is for a gadget to recall your choices, so you do not need to enter data each time you use it.

For instance, you might place an order from Amazon and need to tell it your location and zipcode via speech. In any case, when you've placed an online order once, Amazon will remember those address details and basically request that you confirm it whenever you’ll again be placing an order via speech.

 

Voice User Interface configuration is a promising fresh out of the box field that gives solutions through voice control. With VUI and GUI connected, human-machine communication can be improved and streamlined utilizing input through voice and gestures.

Designers ought to consider the requests that should be handled when they represent the interactive conduct between the user and the system. Examine the information to comprehend where the system has turned out badly. By the day's end, VUI designers need to check and enhance the system persistently.

Need help with VUIs or website design? Get in touch

Contact Us

Topics: NLP, Conversational UI, Speech recognition, Natural Language Processing

Contact Us

Recent Posts

Recent Posts

Most Popular