How to design your voice’s product persona? The good, the bad and the noisy (Part A) - Product of Things
post-template-default,single,single-post,postid-16036,single-format-standard,give-test-mode,give-page,qode-quick-links-1.0,woocommerce-no-js,ajax_fade,page_not_loaded,,qode-title-hidden,qode_grid_1300,qode-content-sidebar-responsive,columns-4,qode-theme-ver-11.0,qode-theme-bridge,wpb-js-composer js-comp-ver-5.2.1,vc_responsive

How to design your voice’s product persona? The good, the bad and the noisy (Part A)


Amazon has recently announced a 20,000 connected devices milestone to its Alexa voice assistant. This brings us rapidly to an IoT connected world, where your customers can talk to and with devices, making a greater product offering.

This is a guest post by Galit Galperin, a voice-product expert. To read more of her writing, visit her blog.


As a product manager, this should get your attention in examining how and if your product has a voice. Voice assistants like Siri, Alexa, Google Assistant, and Cortana are going to change the way we experience and interact with technologies by reducing friction from a large amount of daily task and routines.

 If you think your product vision can fit into this, not just as a feature, but also as a new experience, then you will probably need to start thinking what will be the proper platform to integrate your product’s voice strategy to, and if there is a problem you didn’t consider in your roadmap.

 In order to do this, voice assistants must learn how to deliver exact information and accomplish specific tasks for its users.
As you may know, all the relevant product divisions at the “big four” (AKA Amazon, Google, Microsoft, and Apple) are currently competing to improve the NLU (natural language understanding), NLP (natural language processing) and AI components of their platforms in the hopes of offering a better voice platform; one that can serve users, developers, brands, and all other voice ecosystem stakeholders.

 Did you ever ask yourself what your product would sound like if it had a voice? If you do have that question in mind, this article will help you set the scene for building your product’s voice persona.



So, what is a Voice Product?

 A “product” in the voice world will most likely be a skill or an “action”, an equivalent to the familiar concept of apps. The core differences lie with the user’s visibility – somewhat of a parallel blindness where your user can’t see your product/brand and most likely is occupied in a multitude of parallel actions (cooking, driving etc.) This is quite a leap from a mobile experience in which the user is likely to be watching and interacting with a screen.

When creating an Alexa skill or a Google action, it’s crucial to think about your product’s personality. This would aid in creating an authentic dialog and a more engaging experience for customers.

There is no such thing as a voice user interface without persona “ (Cohen, Giangla and Balogh, Voice Interface Design)



So… how to design your product’s voice persona?

First, let’s understand what’s a persona?

In the voice user interface world, it is defined as a character that is adopted by an author or an actor; Just like a character in a film or a movie. Each persona is different- it’s who you are and what distinguishes your persona from others.

The personality of your VUI (voice user interface) will determine its behavior across the board – how it answers questions, how it offers assistance and how it handles errors (cases in which the machine didn’t understand the user’s intent).

Now imagine the market numbers: 500 million devices that have a google assistant, 50 million smart devices that have Alexa in users’ homes… and counting. Different manufacturers are starting to offer a large variety of smart devices with or without a screen. These are different, fragmented surfaces, but the user’s expectation stays the same for all of them – to have a conversation for each and every surface, for each context.

Persona Design is a crucial part of your product’s voice offering in order to create trust. You should ensure follow some basic rules when building a voice persona, so it will not be broken.



A broken persona equals loss of trust which translates quickly into a loss of users. In cases of no speech detected, or not recognized one, instead of saying “I’m sorry — I didn’t get this” and drop the conversation, you should design it so it will engage with the user, offering them another choice, just as you would in a real person to person conversation:


User: “Play me the startups podcast.”

VUI: “I have a podcast named “startups pitch”, do you like to hear it?”

User: “Uhh, I’m not a fan of this one.”

VUI: “I’m sorry, did you want to look for other podcasts, or will this one work?”


What are the steps and consideration to take when designing a voice persona?

A wide range of surfaces

So, AI needs to move across multiple surfaces effortlessly – users can speak to their TVs, watches, smartphones, cars, and other smart devices. In some cases, the voice will involve a touch screen, in some  — not. Some of these are “hands-free”, some are also “eyes-free”.

For this purpose, we as product managers, need to think not just of the voice itself, but also on sound design such as; (consider: are there any special effects?), interaction design (consider: what happens when the voice assistant is combined with a touchscreen?), what visual content to display and even elements such as typography.


Voice tone

Your potential users can inexplicitly determine a lot of data based on the sound of a voice, specifically experience the brand/product in a particular manner.

Simply put, the user judges your product by its voice and tone.

The user can determine specific characterizations like age, gender, education, social context (my boss, my peer, my family), locality, likability, intelligence and so on.

When defining your product’s voice persona, you need to define and reflect on all of these matters, so it will be recognizable.  

You can decide if to use TTS (text to speech) option, or custom recording based on the content you intend to share (for example: is the content dynamic?). Take into consideration that TTS is probably more scalable and less expensive than a dedicated recording, but not as personal and unique.


Understanding the brand/product

Review the brand style guide, the different touch points with customers, the product experience, the social media reflection and so on. Your voice product should reflect all these as well.

Let’s take the banking industry as an example, think about the different voice persona from one bank to another. The first aims to reach the elderly, and the other is defined as a digital bank for millennials. It would be safe to assume that they will have not just different voice persona, a tone of voice and re-prompts, but also different product characters and features.


Understanding your audience

A product manager has to know its product’s audience. Nothing new here.

Who is going to use your skill/action? What is the context? Look at the customer journey, demographics and frequency of engagement – these will determine your voice prompt. If your customer engages frequently, then your prompts should be shorter and vice versa.  You can use this nice tool to create one or more of your customer personas if needed.


Understand the task

What is the customer trying to achieve? What are your product goals? How can the customer achieve these? Try to break it down into steps. Sometimes it might require creating more than one task or one persona to fulfill it.

If we take the previous example, the skill for the elderly might have two different product features for two different personas — where one is stating only the current balance, and the other allows a more complex task for advanced users, like a bank transfer to family members.

At the same time, the millennials bank can offer the same product features but will sound completely different, and will use different language and prompts, or even offer a more advanced feature, with a different flow designed for the specific segment. As a product manager, you need to be able to distinguish between the different tasks and use cases.


Did you find this post interesting? Make sure to read its second part next week!



No Comments

Post A Comment