Voice Assistant in Mobile Apps: My Jio

Welcome to a new blog series by Slang Labs, called "Voice Assistant in Mobile Apps". Here we tear down the voice assistant and search functionalities that have been added by businesses to their mobile applications and discuss them in detail. In this edition, we breakdown the voice search feature in 'MyJio' Android app.

We believe that it's essential to recognise the trendsetters and show how they are adding voice assistants inside their applications and its result. This analysis will also help developers understand voice search's usage and the importance of voice assistants in applications. We have already broken down 'Gaana', 'YouTube' and 'Paytm Travel'.

Jio launched Hello Jio last year inside their 'MyJio' app. One can access this assistant by clicking the mic button in the search bar. 'Hello Jio' as of today works in English and Hindi. Most of the actions that it can take today through voice search are inside the MyJio app.

There is a lot to break down, so this might be a long one. Buckle up!

Visual Breakdown

New User eXperience / Onboarding people on voice

After giving all the permissions (your usual mundane traditional onboarding), we land on the main screen of the MyJio app. Jio doesn't nudge individually for the mic button. Surprisingly, it does show a nudge for the QR scanner for some reason. Upon clicking the mic button in the search bar, we get to see some insights. Jio has put in the effort to add a voice assistant inside their app.

Starting with Language Selection

Voice journey of the user begins with a language selection screen. As of today, Hello Jio only supports two languages, namely English and Hindi, which a user has to select to get started with their Voice Assistant.

We recommend doing this as well for voice on-boarding. With our latest update, Slang also starts the user journey by displaying users, a language selection screen. Upon selecting a language, there is auditory feedback confirming the selected language.

Coach Marks

Jio shows an image explaining all the UI elements on the voice assistant screen in one go. In our opinion, this is not the right way of introducing a new feature to the users who are coming online for the first time. You have to click on the cross button to move to the next screen.

Mic permission

Jio speaks out "What can I do for you?" and simultaneously asks the user to give mic permission.

This action tells the user why the permission is required. In our user research, we also found out that it's better to tell the users why we need permission. Today, Slang does so by first speaking out why we need the mic permission and then showing the permission dialogue. We eventually want to move towards asking and simultaneously demanding the mic permission.

UI Breakdown

Unlike Slang, which is an overlay on the existing app screen, Jio's voice assistant is an app section by itself.  It tries to give the user the answer on the same screen or take them to the relevant page.

Jio has made sure to show the users (in the language they selected during onboarding) what they can ask the voice assistant to do.

These hints are shown even when a mic button is pressed. These hints help the user see what they can ask through voice search and which format they should ask.

Today, the tasks are more or less limited to in-app features of the MyJio app. But, Jio has added apps like JioSaavn and JioCinema inside the MyJio app, which increases the spectrum of voice search functionality.

The voice assistant screen consists of 5 elements:

Mic Button

‍Placed front and centre on the screen is the mic button, which users can tap to speak to the HelloJio assistant.

Light/Dark Mode

There are two display modes available, namely, Light and Dark mode. This icon on the top left corner of the app is for this functionality.

Keyboard Icon

Users can enter the query via the keyboard, alternative to speaking it out.

Language Toggle

This button allows users to toggle between Hindi and English.

Listening Mode

On clicking the mic button, multicoloured waves pop up and move, which indicates listening. The utterance spoken by the user is displayed on the screen in bold on top of the screen. The screen also shows the language which the app is currently listening in.

Functional Breakdown

Voice to Action

Jio Assistant uses voice to action model to take the user to the most relevant page based on the input.

Even if the user specifies some data for the desired intent, Jio's assistant doesn’t fill these details on the destination page.

If a user says, 'Recharge my account for Rs. 399', they still end up at the same recharge page.

In Slang's case, we take the user to the closest possible step based on the data provided by the user. In the above example, Slang would take the user to the payment method page, bypassing the pack selection screen, as we already know the pack user wants.

Onscreen responses

Just like chatbots, Jio assistant also tries to answer some questions directly on the same screen. There are no explicit heuristics that clarifies when the assistant replies in the same screen and when it takes you to the relevant page.

Handling Missing Data

Jio handles missing data by showing up the possible options on the screen. The user has to fill the missing data by selecting the option.

Slang's framework auto-prompts the user to collect the mandatory data via voice, without having the developer to write a single line of code.

Editing query:

Jio allows the user to edit their utterance by showing an 'edit' option next to the spoken phrase. This, although only happens via touch.

What’s still missing?

Although we covered most of the changes we recommend earlier in the blog, these are still some more changes that we think can significantly improve Jio's voice experience:

Better NLP capabilities‍

Jio can improve the NLP capabilities of the voice search to make the overall voice experience user friendly.

Hindi capabilities

More vernacular languages

Jio, which is used all over India, doesn't support any other language apart from Hindi.

Better action handlers

Even though the Jio assistant hints tell the user that they can ask it to play a song, the user is taken to google play store to download Jio Saavn, even though you can play the music via touch inside the MyJio app.

Slang allows you to add voice search in multiple Indian languages in a fraction of the time with inbuilt analytics to gauge how users are using it. It also gives businesses the ability to add more voice-enabled functions quickly to the app.

‍