Voice personas
Oct 07, 2021
7 MIN READ

How to Make Your Voice Assistant Likable and Relatable

Why should your voice assistant be likable and relatable? Why does it need to have a personality that matches your brand’s tone and values? As humans, we tend to form attachments, and voice is one of our most human characteristics. It makes sense that voice assistant users often form attachments to their digital counterparts and begin to see them as friends and companions. 

For brands, these attachments can translate to brand loyalty and affinity that goes beyond just the voice user interface. Taking the time to create a voice assistant that truly reflects your brand values and one that delights the user, is time invested in reducing churn and building a future customer base.

Choosing the tone, voice, personality, and level of humor will help determine how your brand is perceived and what your users will begin to expect from you. It can also be the difference between a user becoming frustrated at an error and abandoning the voice AI or being understanding and trying again. The more of a connection that forms between the user and the voice assistant, the more dynamic, rewarding, and meaningful the experience will be for your customers and the higher their level of satisfaction with your brand. 

According to Statista, the global voice recognition market size is forecast to grow from $10.7 billion in 2020 to $27.16 billion by 2026. With more brands investing in voice AI, companies need to be thinking about how a custom voice assistant can help differentiate them and their products and focus on creating one that truly speaks to their audience’s needs, wants, and expectations. Settling for the same voice assistant as your competitor or accepting a voice experience that doesn’t engage your users will make competition in crowded markets harder. 

Instead, infuse your branded voice assistant with these 3 qualities:

  1. Promote anthropomorphism
  2. Add personality 
  3. Incorporate empathy

1. Give your voice AI human-like qualities

It’s easier for users to connect with a voice assistant whose character and manner of speaking are familiar and recognizable. Unless your brand is specifically going for robotic styling to convey futuristic and intelligent aspects, you’ll want to make sure your voice assistant has human-like qualities, including: 

  • Be conversational 
  • Respond with good grammar
  • Express appropriate intelligence
  • Determine the best response length

Instead of blunt answers to queries, it’s important to add some keywords and phrases in between to keep the voice assistant conversational and convey empathy to the user. It also establishes a level of friendliness and detail that users will appreciate. 

For example, if a user asks a voice assistant to make a reservation, the voice AI could follow up with the question, “What is the name of the guest?” The response does get the point across and moves the conversation forward but lacks helpfulness and likability. Instead, the voice assistant could respond with “Sure! I can help you with that. Can you tell me your name?” and instantly put the user at ease and start forming a connection with them. 

“Never assume that you have enough utterances. Don’t just think about “How would I ask this?” Because if you ask your friend, he will probably ask it very differently.”

Mihai Antonescu
Senior Engineer
Mercedes-Benz R&D North America
@MBRDNA

It’s also vitally important to be grammatically correct. It may seem like a small detail, but ungrammatical responses create an instant reaction in the user that something is wrong with the voice assistant. It could be that they think the voice assistant lacks intelligence, is less than human, or is poorly designed. Either way, hearing something simple like “I have two orders of french fry for you” takes them out of the moment of interaction, and the user may be skeptical about the voice assistant’s capabilities while forming a subconscious dislike for it. 

Another factor to consider is the level of intelligence your brand wants to convey to the user. Voice assistants are a complex technology capable of calculating, personalizing, and cross-referencing information at a speed most humans aren’t capable of. Some users are comfortable with their voice assistant being smarter than them. Others feel ill at ease at the prospect of technology being more intelligent. Brands should evaluate their users’ wants, needs, and expectations and even incorporate this question into user testing to determine what the right level is for their audience. 

Finally, there’s the length of responses. Too short of responses can make users feel uncomfortable with the abruptness, while too long of a response can make them feel annoyed at the delay. Either way, the negative feelings remove the connection between the user and the voice assistant. 

According to a study by Apple researchers, 70% of people prefer more conversational responses to terse answers. People who identified as chatty preferred chatty responses, while those who identified as non-chatty preferred short responses. This signifies the importance of user testing and knowing your audience to determine the length of responses that will make the best connection between the user and the voice assistant. 

70% of people prefer more conversational responses to terse answers.

Apple Researchers

2. Infuse personality into your voice assistant

One of the fastest ways to get users to like and relate to a voice assistant is to have them connect with the voice AI’s personality. A personality connection forms a deep bond between the user and voice assistant where the user attributes human-like qualities to the voice assistant and considers them more of a friend than just an AI, which will lead to increased usage. Brands will need to create a voice persona, identifying who will be using the voice assistant, what their needs are, and where they will be using it. 

Here are 4 elements to consider for creating a voice assistant with personality: 

  1. Voice
  2. Tone
  3. Word choice
  4. Humor

A great starting point is with the actual sound of the voice assistant’s TTS response. This is what the user will be hearing and interacting with, sometimes many times a day. The user will make either conscious or subconscious decisions about how they feel about the voice assistant based on the gender, pitch, and tone selected. It’s a good practice to spend a lot of your user testing on the voice to gather what appeals to your audience and what doesn’t. You’ll also want to consider what your users’ geographic locations and cultures are. Certain areas are more sensitive to tone, and levels of respect and formalities will vary by culture. 

“The intersection of a social issue and technology is always going to be rough. The best way to be prepared for it is to have a good answer for why you chose the gender you did.”

Lauren Golembiewski
CEO and Co-founder
Voxable
@LaurenGolem

Phrases and word choice are also key elements when creating a likable and relatable voice assistant, and it largely depends on the purpose of the voice assistant—if it’s for fun or functionality. Tasks such as setting a timer, sending an email, turning on the air, or checking the weather are things people want to do quickly, efficiently, and accurately with no room for small talk. Other queries like telling a joke, telling a story, or random questions leave more room to delight the users with unexpected responses or even humor.

Humor is one aspect that brands need to be especially careful with, though. What some people may find funny, others may not appreciate. User testing comes into play again here with finding a good balance between super cheeky and boring responses. The sweet spot is usually somewhere in the middle. When appropriate, humor can be a great asset for marketing and to form a strong connection with the user. 

However, if the voice assistant is for healthcare, business, or education, there may never be a time when humor is appropriate, which is why evaluating the purpose of the voice assistant is vital. If the user base is mostly children, then the more humor, the better. If it’s for surgery, then humor should probably be avoided. 

“A VUI, like any technology, isn’t perfect. Designing humorous or witty quips and come-backs are valuable in keeping the user smiling when the system can’t deliver the desired information or action.”

Rachel Batish
VP of Product
Audioburst
@rachelbatish

It’s also important to note that personality should be unique and based on what works best for the brand. Not all voice assistants have to be the same, and in fact, brands should be encouraged to explore different styles. Just as humans have a variety of personalities, so should voice assistants, as it makes for a more interesting ecosystem and user experience. A personality that truly represents the brand’s tone and values can be a differentiator in the quickly expanding voice market. 

Developing a personality also takes time. It’s a challenge, and you may not get it right the first time. User testing is essential to see how users respond to tone, voice, and humor and if it matches their expectations. A voice assistant’s personality can be iterated over time and fine-tuned into one that users will form connections with.

3. Incorporate empathy into the voice experience

Empathy is one of the most human-like qualities out there and a great way to gain a user’s trust and respect. Nothing makes a user more frustrated than getting an inappropriate or inaccurate response from a voice interface. However, if that response reflects empathy and understanding, the user may be more forgiving—as they might be with a human agent. In the near future, as voice AI advances and natural language understanding (NLU) evolves, voice assistants will be able to adjust responses based on the tone and mood of the user. For now, developers must rely on some common situational parameters to design the best voice experience possible.

“As designers, we must anticipate and predict intent wherever possible. The more we can enable the tech to intuit intent and anticipate requests, the more frictionless and empathetic our experiences will become.”

Will Hall
Chief Creative Officer
RAIN
@RainAgency

While empathy is great for everyday usage of voice assistants, it is even more valuable for situations like voice-enabled call centersdrive-thrus, and kiosks where customer satisfaction is of the utmost importance. Voice AI could potentially change responses if they detect a frustrated or angry tone or connect them with the right human representative trained to handle such situations. 

It’s also important for voice assistants to know when not to respond. While sensing frustration based on tone is still to come, voice assistants can be designed to sense frustration based on words. For example, if a user tells a voice assistant to shut up, then the voice AI should know to stay quiet and that nothing said could improve the situation. 

The more a user likes and relates to a voice assistant, the more likely they are to use the voice assistant, recommend it to others, and be understanding if there are simple errors. Using elements such as grammar, intelligence, personality, voice, tone, word choice, humor, and empathy can increase the connection between the user and the voice assistant. It’s essential to spend adequate time on research and user testing to determine what personality or style best serves your users. If a user doesn’t have a connection with a voice assistant, like, or relate to it, they may abandon it for a voice-enabled competitor. 

At SoundHound Inc., we have all the tools and expertise needed to create custom voice assistants and a consistent brand voice. Explore Houndify’s independent voice AI platform at Houndify.com and register for a free account.

Kristen is a content writer with a passion for storytelling and marketing. When she’s not writing, she’s hiking, reading, and spending time with her nieces and nephew.

Interested in Learning More?

Subscribe today to stay informed and get regular updates from SoundHound Inc.

Subscription Form Horizontal