• Henal Jasani

Future of Google Voice: Connecting with Customers in the Age of Voice.



Voice technology has been developing quite rapidly in the past couple of years. We have more voice-enabled applications and devices around us than ever before. Even our mobile phones come with a voice assistant that can help us with our tasks and even hold a conversation with us.

25% of US adults own a smart speaker, and 30% of them are active voice users.

But these days, you can find voice on different devices such as watches, headphones, and more.


Most US adults make use of a smart speaker to reduce on-screen time. Parents mainly buy a

smart speaker for this reason. Their kids don’t have to be glued to a screen to access the

internet.


Speech recognition systems are measured for accuracy with a word error rate which is currently at 4.9%, lower than the Human word error rate (5.1%).


The various opportunities and challenges in the voice UX world are quite similar to the ones in the chatbot world, like picking the right use case and making it work well.


People tend to use voice daily and not just when it is thrown at them via automated assistants. It caters to all their needs, such as ease, convenience, and accessibility.


How do Companies use Voice Technology?


Picking the Right Use Case


Before any company implements voice technology, they can ask themselves -


● Is it reducing the amount of effort the users have to put in?

● Is it allowing users to complete the full journey by voice?


As a company, you must look for opportunities to make it easier for the customers to interact

with you via voice tech. For complex functions, users may open up a chatbot or browse the

computer. However, you can provide them with a voice feature for simple tasks.


Make it Work Well


For any company to implement a voice model, they need to make it work well after picking the right use case. There are different areas that the company can focus on, ensuring that voice can be the right move.


When a user visits your application for the first time, what do you present them with? What

obstacles will they face, or will you make it easy for all users?

You can get better at this by experimenting and analyzing the data of users interacting for the first time. You can also invest in Natural Language Understanding (NLU).

If done well, users will learn how to interact with your bot and understand what works and

what doesn’t. Your bot can prompt them with a primary objective that most users tend to look for in your application. This way, you can set it for failures also.


Most conversational designers consider providing menus a part of NLU and not a basic human conversation. But the reality is that even in real life, we do offer choices and menus in everyday conversations, and including menus makes it easier for customers to respond correctly to the bot.


Context Impersonalization


Personalization is a huge form factor for customers, but you cannot provide this feature when

you don’t have enough data lying around. In such cases, you can generalize the content for your customers. This is where context impersonalization comes in.


If the bot can recognize context, it can open up with relevant questions, and you may begin a conversation on the right path. You may not need an entire set of data about the customer, just a small piece of information to begin with, such as a recent order or recent trip.


Invest and Answer the Right Questions


Companies need to invest in NLU so that the bot can understand and recognize the context.

With visual aid or UI, the bot can answer multiple questions with just one card, but it is much

more difficult with voice. If the bot does not provide relevant answers, the customers may get confused or find it hard to process all the data.


The message for companies looking to use voice skills are -

● Keep it simple and do them well.

● Lean into convenience.

● Remove obstacles.

● Use user Data.


Final Thoughts


Companies can work on their own voice models, but this can be quite time consuming and

difficult. Many companies also tend to adopt existing voice models. Google Voice is one such model that you can integrate into your application and have it trained specifically for your company.


If you require more assistance in voice technology, you can visit https://design.google/.

© 2017 by Chatbots Conferene TM.