Future Voice Branding: What AI Agents Bring to the Future of Brands?

This article features content originally published in "Design Mind," a design journal operated by frog, and is provided by Dentsu Inc. BX Creative Center Noriaki Okada .

"Hey Siri, can you tell me a story?"

Last night, my children asked Apple's voice assistant "Siri" to read them a bedtime story. Honestly, it was quite good—a perfectly short tale with an interesting plot.The children seemed to enjoy it immensely and begged for another story. Siri then told them a story about a cat. My children are the type who listen intently to people's stories. This is probably related to the fact that I often read to them before bed. Perhaps because of this, I also noticed a rather disappointing fact. When the second story ended, I was startled to realize how flat and monotonous Siri's voice sounded.

I think that's why I couldn't really connect emotionally with Siri's stories. After the children drifted off to dreamland, I started thinking about the importance of voice tone. Whether you feel a heartfelt connection depends less on the story's content itself and more on how it's told.

Voice tone has long been a crucial element of brand expression. As generative AI evolves from disparate tools into real-time, conversational multimodal interfaces, its importance will only grow. Voice tone is the key to the future of audio branding, beyond sound logos and brand theme songs.

※1 Multimodal = Technology and methods that process information by combining multiple types (modes) such as text, video, and audio.

AI agents in a new world where conversational interfaces are mainstream

Brands are seeking to break free from traditional customer engagement tools as dynamic, real-time, interactive experiences With the emergence of generative AI, brands are moving beyond traditional customer engagement tools. The advent of generative search engines is shifting attention away from brand websites, marking a transition from Search Engine Optimization (SEO) to Generative Search Optimization (GEO).

Numerous apps will eventually be replaced by AI agents (AI systems capable of autonomously executing tasks and making context-aware decisions and actions) that interact with all apps on behalf of users. Soon, there will be no need to operate multiple platforms and channels separately.

AI agents will consolidate these fragmented elements into one unified system. AI agents will invisible assistants , seamlessly handling countless tasks across various channels and platforms. Diving into this future isn't difficult. However, it's certain that physical experiences will become even more important in such a future—but we'll save that discussion for another time.

We will eventually be freed from daily tedious administrative tasks. Planning a weekend getaway while exercising on a treadmill will be a breeze. Before you've even burned off the extra calories from breakfast, your AI agent will pick a destination, arrange transportation and hotels, order books tailored to your tastes, and even make restaurant reservations. It would be even better if it could handle the exercise for you too.

That said, if you sprain your ankle while running on the treadmill, you can ask your AI agent to order an ankle brace. Moreover, beyond such administrative tasks, when you need emotional support, your AI assistant will respond with a warm, human-like tone of voice. Unlike Siri's monotone voice reading you a bedtime story, the voice of future generative AI will be warm and emotionally resonant.

"Hey AI agent, can you tell what kind of mood I'm in today?"

In a 2024 TED Talk Microsoft AI CEO Mustafa Suleyman argued that generative AI must be viewed as far more than just a tool or platform. He called for treating generative AI as a new "digital species." Building on this idea, he suggested that generative AI requires respectful treatment and time to nurture.

"When we build AI, we can and must reflect all the good things, the things we love, the best human traits—empathy, kindness, curiosity, creativity." (Mustafa Suleyman)

Following this logic, AI agents can become more than just a voice. The shift to natural language enables more intimate interactions with customers. By learning user preferences and habits, AI agents will increasingly understand the context and emotions of each moment. To be clear, the ability to understand emotions is distinct from general intelligence; I am not suggesting AI agents become animated or possess consciousness in any form.AI agents will still be remarkably intelligent digital ecosystems, but as their knowledge about users accumulates, they will learn how to craft voices that feel intimate to those users.

Moreover, voice isn't just one-sided in these interactions. AI agents are also listening to the user's voice. This means they can convey what the user wants to know in a tone of voice appropriate to the prompt.

For example, if you ask whether you can play golf this afternoon or what the weather forecast is, it might respond with an enthusiastic tone if it's sunny, or a disappointed tone if it's rainy and windy. Taking it a step further, it should eventually be able to adapt its interactions to each individual family member. Returning to the earlier example of telling children a story about a cat, you can imagine it choosing a more playful voice tone when portraying the cat character.

"Eventually, everything will be possible through conversational interfaces," (Mustafa Suleyman)

Voice has immense influence, making it a gateway to broadly engage senses beyond hearing—such as sight, touch, and smell. A new world where internet connectivity is commonplace AI agents may surpass our imagination. They hold the potential to become far more than mere tools for making reservations or ordering goods and services.

Recently, there are warnings about a sharp increase in people feeling lonely and experiencing related health issues. It wouldn't be surprising if the number of such individuals decreased with the spread of AI agents. However, regarding the role of brands, loneliness might not disappear. For example, if a user's AI agent only interacts with a brand's AI agent, isn't that essentially the same as no connection between the user themselves and the brand?

The Current Position of Brands

The word "brand" carries the meaning of "burn," originally referring to the practice of branding livestock to indicate ownership. Over time, it came to denote the mark placed on goods to indicate manufacturer or origin. Today, the term "brand" represents a crucial element for an organization's presence. Yet, in the new world guided by AI agents, the role of brands risks diminishing.

Whether developing their own AI agents or optimizing those of others, it will become essential for brands to create a distinctive voice layer that stands out as much as possible. This can be thought of like a layered birthday cake.

The top layer is the candles and toppings (the brand layer). The next layer is the icing (the emotional layer built to match the situation and each customer's preferences). The rest of the cake is the LLM (Large Language Model) layer, the engine driving the voice.

Considering all this, one question arises: How should we determine a voice that truly represents the brand?

"A brand must define its unique voice that aligns with its mission, values, and target audience. The brand voice must reflect the tone and style that best suits the brand's personality, identity, and communication goals. It must also leave a consistently recognizable and memorable impression on customers across all channels and platforms."

The above passage was generated by AI. While not incorrect, it's hardly creative writing. The answer to how to define a brand's voice layer likely already exists within existing brand expressions. However, when shifting focus from brand guidelines to a branded voice, that is merely the starting point.

It must creatively articulate the brand's story in alignment with the ideal sound of that narrative (and, furthermore, avoid the need for massive expenditures to license celebrity voices). Various data sources—anonymized personal information, extracted and analyzed public data, research findings, and even synthetic data—become the secret weapon for crafting the brand voice.

For example, a consumer goods brand with a famous founder could build its brand voice using the founder's philosophy and recordings, alongside various phrases that capture the brand's personality. A global sports brand might do well to prepare multiple voices for different sports, utilizing the voices of particularly influential ambassadors or campaign voices, rather than focusing on just one voice.For financial services brands, recording and using the voices of customer service staff while paying attention to regional characteristics and needs is another viable approach.

Capturing the Future Sound Echoing from Afar

Among the world's most forward-thinking organizations, especially those with distinctive brands, some have already begun exploring unique brand voices and conversational interfaces for a new era. The "brand voice" they discover will shape like other parts of the customer experience (CX) strategy strategy. However, unlike other parts of the CX strategy, the brand voice is the most emotionally resonant asset in the brand portfolio and must be carefully protected and nurtured. Next time my daughters read to me, I hope they'll share heartwarming, delightful cat stories using a voice nurtured in just such a way.