For most of the last two decades, building a brand meant winning the visual layer: the logo, the palette, the typography, the feed. The eye was the organ brand strategy was designed for. Sound was an afterthought — a jingle here, a sonic logo there — too subjective and too hard to control to plan around seriously.

That hierarchy is quietly breaking down. AI-based voice interfaces are giving sound, tone, and presence a new seriousness. At the same time, the graphical surface itself is losing some of its grip, as natural user interfaces — voice, gesture, situational behavior — take on more of the interaction. Feeds keep getting louder while attention keeps getting thinner; researchers now describe a genuine digital fatigue, backed by recent studies, from the sheer volume of images competing for the same shrinking pool of attention. Something has to give, and increasingly what gives is the primacy of the image itself.

Where does attention go instead? Toward voice. Toward posture. Toward whatever a brand’s aura actually is when you strip the logo away. And, increasingly, toward physical rooms — toward the old, almost unfashionable idea that people remember what they experience with their bodies far longer than what they scroll past.

We got a concrete lesson in this with the Audi Club of Progress, an immersive live marketing tour we built across five German cities, pairing test drives with a multisensory dinner format. It wasn’t a trade show booth or a showroom. It was a room designed so that people could meet a car the way you meet a person — through presence, not through a spec sheet. One part of that installation gave the Audi A6 e-tron its own AI-developed voice, deliberately built to sound less like a product and more like a personality: warm, present, a little charming. A visitor’s reaction afterward said more than any brand tracking study could: she said the voice was so likeable she fell for it on the spot. That’s not a comment about audio quality. It’s a comment about what happens when an experience is designed as an encounter instead of a channel.

That distinction — encounter versus channel — is really the whole argument. A channel broadcasts the same thing to everyone and hopes enough of it lands. An encounter is built to respond to the person in front of it. And this is precisely where AI turns out to be useful in ways that have nothing to do with generating more content faster.

Inside a live event, AI can adjust pacing, content, and even the arc of an evening in real time, based on who’s actually in the room, without turning the experience impersonal — if anything, the opposite happens. Inside a physical space, AI can generate and vary projections, lighting, and soundscapes so a showroom or pop-up responds to who’s standing in it rather than running the same loop regardless of audience. None of this replaces craft or design judgment. It compresses the distance between concept and iteration, so ideas that used to take a full production cycle to test can be prototyped and refined in days.

This is also why we built think moto +AURA as a dedicated unit rather than folding sound and space into whatever team had bandwidth that quarter: audio, voice, and immersive design deserve the same strategic seriousness as visual identity, not a smaller budget line labeled “experiential extras.” Treating body, sound, and presence as an afterthought is exactly how brands end up with a beautiful logo and a forgettable room.

The economic logic underneath all of this is straightforward, even if it’s easy to miss while everyone is busy talking about content velocity. Generative AI is making an enormous share of brand communication infinitely reproducible — more copy, more images, more variations, at a cost approaching zero. Anything that can be copied that easily eventually stops being a differentiator. What doesn’t get cheaper is a specific place, on a specific night, that a specific group of people actually walked through. Scarcity is shifting from information to encounter.

That’s a strange, almost counterintuitive consequence of an AI-saturated market: the more content there is, the more valuable a well-designed room becomes. Brands that treat live, spatial, and sensory experience as a strategic layer — not a seasonal marketing stunt — are quietly building something that doesn’t show up in a screenshot and can’t be scraped by a competitor’s model. Everyone else is going to keep optimizing the feed a little harder, for a little longer, until they notice that the feed itself has stopped being where trust gets built.

This essay is a shorter companion to the full article, published on the think moto blog: Hear It, Feel It, Live It: Event, Experiential and Immersive Design in the Age of AI.

Brands Are Starting to Live in Rooms Again, Not Just in Feeds was originally published in SPHERICAL THINKING on Medium, where people are continuing the conversation by highlighting and responding to this story.

Ask most marketing leaders whether their brand is ready for AI, and you’ll get a confident yes. Ask what that readiness actually consists of, and the answer usually gets vague fast: a chatbot here, some AI-assisted content there, maybe a pilot with an AI customer service assistant. All of that looks like progress. None of it answers the real question.

The real question isn’t which AI tools a brand has adopted. It’s whether those tools — and the systems built on top of them — actually know what’s right for the brand when they act on its behalf.

That distinction matters more than it sounds. Individual AI tools are exceptionally good at one thing: scaling output. A brand can now produce ten times the content, respond to customers instantly, and generate campaign variations in minutes. What none of these tools solve on their own is whether that scaled output stays coherent with who the brand actually is.

This is where most AI adoption in brand marketing quietly breaks down. One tool writes social copy. Another handles service queries. A third drafts campaign concepts. A fourth reviews content for compliance. Each one works from its own prompt, its own data, its own assumptions about the brand. Individually, each tool performs well. Together, they don’t produce a stronger brand — they produce noise wearing the brand’s colors.

An AI-ready brand looks structurally different. It has a clearly defined, machine-readable brand foundation — values, tone of voice, positioning, guardrails — that every AI system draws from, rather than each tool inventing its own interpretation. Brand decisions aren’t delegated to whichever model happens to be answering; they’re derived from a shared source of truth. In practice, this shared foundation is sometimes called a brand core: a structured, queryable version of what a brand stands for, written so that machines can act on it as reliably as trained humans do.

So how does a brand find out where it actually stands? A useful gut check runs across five areas.

Governance comes first: does anyone in the organization know, concretely, what AI systems are and aren’t allowed to say or do on the brand’s behalf? Without an explicit answer, the model decides by default — and “by default” is rarely the same as “on brand.”

Brand memory and data quality come next. A beautifully designed brand guideline PDF sitting in a shared drive is fully readable by a human designer and functionally invisible to a language model. If AI systems can’t retrieve brand knowledge in a structured form, they will improvise — politely, fluently, and often incorrectly.

Sourcing and authenticity raise a quieter but growing risk: content generated at scale, without brand oversight, doesn’t just fail occasionally — it fails systematically and invisibly, until the drift becomes obvious to customers before it’s obvious internally. The moment an AI system starts talking to customers directly, it has effectively become a branded agent, whether anyone designed it to be one or not — and it deserves the same scrutiny as a human-facing brand touchpoint.

Compliance and transparency are no longer side issues, particularly for European brands. Where brand data is hosted, which models process it, and whether customers are told they’re speaking with an AI are increasingly not just legal questions but brand-trust questions.

Measurability closes the loop. Without a way to measure whether AI-driven brand interactions are strengthening or eroding brand equity, “AI-ready” remains a slogan rather than a verifiable state.

None of this is really about technology adoption speed. Plenty of brands that consider themselves behind on AI are, structurally, in better shape than brands that have deployed a dozen AI tools without ever building the foundation underneath them. The tools were never the hard part. The hard part — as it always has been in brand management — is deciding, clearly and in writing, what the brand actually stands for, and making that decision usable by systems that don’t share human judgment.

That’s a strategy problem before it’s a technology problem. It’s also, increasingly, the dividing line between brands that will look coherent across a thousand AI-mediated interactions a year from now, and brands that will look like they’re speaking in a dozen different voices at once — all of them technically “on-brand,” none of them quite right.

None of this is solved by running the assessment once. The five areas above are a starting point, not a finish line — the more useful version of this check is continuous: which parts of the brand foundation are machine-readable today, which AI systems actually draw on them, and where does governance still fall short. That’s less a checklist than an ongoing discipline.

This essay is a shorter companion to the full article and ten-question self-assessment, published on the think moto blog: Is Your Brand AI-Ready? 10 Questions Every CMO Should Be Able to Answer.

Is Your Brand AI-Ready? Most Aren’t — and It’s Not About the Tools was originally published in SPHERICAL THINKING on Medium, where people are continuing the conversation by highlighting and responding to this story.

Imagine a mid-sized company that has spent years building its brand.

Clear positioning. Consistent design. A brand voice that works reliably across campaigns, websites, sales materials and corporate communications.

Then AI arrives.

With it comes the pressure to produce faster, more, and cheaper. Marketing starts using AI to generate copy. Design experiments with image and layout tools. Customer service builds a chatbot. Research teams use AI to summarize interviews and market data. Legal and compliance begin testing automated review workflows.

Each initiative makes sense on its own. Each solves a real problem.

And three months later, every department is producing “brand-consistent” output — in five different tones of voice, with three different value propositions, and a chatbot that answers customer questions as if it belongs to a different company entirely.

The tools work. The brand doesn’t.

This is not an edge case. It is the structural problem behind the AI boom in branding. Companies are not implementing one AI-enabled brand system. They are building, buying and connecting separate AI solutions for separate tasks , often without a shared understanding of what the brand actually is.

That’s where fragmentation begins.

The fragmentation nobody talks about

The market is solving the wrong problem. DAM systems manage assets. Content AI tools generate copy. Design systems structure visual language. AI assistants handle conversations. Approval tools organize workflows. Each solves a real partial problem. None solves the actual problem.

The actual problem is not the lack of AI tools. The actual problem is that most of these tools do not draw from the same source.

One tool for analysis. Another for content. A third for approvals. A chatbot trained on one set of documents. A copy assistant prompted with another. A design check based on visual rules, but disconnected from strategy, voice and meaning. Every new tool that solves one problem creates three new inconsistencies. The result is more production, but less control.

So the central question is not: which AI tool should we use?

It is: what shared brand foundation should every AI-supported brand decision draw from?

Three layers. One source.

Brand leadership in the age of AI requires three capabilities — simultaneously, and from the same source:

Intelligence: Understanding your brand’s position, gaps, and competitive context. Not just data. Judgment.

Production: Translating brand knowledge into outputs. Copy, assets, AI assistants. But conditioned by the brand core — not by a generic prompt.

Governance: Actively securing consistency before content goes out. Not archiving what’s already approved. Reviewing what hasn’t been yet.

Governance without Intelligence is blind rule enforcement. Production without Governance is uncontrolled scaling. Intelligence without Production remains analysis without impact. Separated, they produce more output and less brand.

Why method is the real differentiator

There’s one more thing the market rarely discusses: what knowledge base the system is actually built on.

Most branding AI tools are built on generic foundations — prompt templates, style guides uploaded as PDFs, rulebooks that capture the surface of a brand but not its logic. The system knows what the brand looks like. It doesn’t know what the brand means.

Spherical Brand AI is built on two proprietary frameworks: the Spherical Brand Framework and the Brand BIOS® Model. The system doesn’t just know what the brand communicates — it knows why. That’s not a feature. That’s an architecture decision.

And it’s why no competitor can replicate it by adjusting a prompt.

The full argument, including the three-layer architecture, diagnostic questions for CMOs, and FAQ, is on the think moto blog:

Why Single AI Tools Don’t Solve the Brand Problem — And What Does

BrandBIOS™ — das think moto Markenmodell.

Your Brand Is Being Automated in Pieces. That’s the Problem. was originally published in SPHERICAL THINKING on Medium, where people are continuing the conversation by highlighting and responding to this story.

Einen Einblick ins Buch gibt es unter Brandedinteractions.de und natürlich auf der Verlagsseite zum Buch.

Messaging-based chatbots often need to share the user’s attention with other chatbots. It is quite possible that a user has several virtual chatbot contacts in his messenger in addition to his human contacts. The user will “ping” them as much as his human contacts:

“Hey Anna, I have to go to London tomorrow for a few days, can you suggest a flight?” “Sure! Do you like to fly alone or for how many adults and children should I search?”

In such multi-bot-environments it’s crucial that the brand remains distinguishable and has a strong, independent character that responds adequately to the user’s context. Same goes for voice assistants, as we see a growing number of specialised services coming up (see our overview of current voice assistants here (German only)).

Here are five things to consider, when defining a chatbots personality:

More than in any other medium, in chatbots the brand personality IS the user experience. How do we encounter the brand in a certain chatbot? As a friend? An assistant? A Challenger? A trainer? At think moto we usually derive the chatbots personality from your brands archetypical role in the market, which it has to relate to, although in many cases it will not be identical with the brands over-all personality.

Also, not every personality fits every task. Different personality facets should be emphasized depending on the situation and application context. Just as a service employee needs different soft skills and character traits than a salesperson, the chatbot that supports the user in service tasks must tick differently than the shopping chatbot.

When we define a chatbots personality, we should distinguish character and behavior. The character remains the same, but over time its behaviour can adapt to the user in a corridor to be defined by the brand and thus contribute to building up a lasting relationship.

A voice assistants or messaging bots personality must match his/her knowledge and skills. What topics and content does the bot feel confident about? Therefore, at the beginning of a project we define a thematic map on which we classify knowledge domains and sort them according to specialist knowledge, general knowledge and “un” knowledge.

Suspension of Disbelief, a term coined by the English poet Samuel Coleridge for defining the readers sacrifice of logic for the sake of enjoyment, is an important aspect here, too. If the user doesn’t get an answer to a question that fits the chatbot, it feels like running against the wireframe grid in a VR environment marking the end of the virtual world. (Well, yes, at think moto we also define VR/AR design pattern for brands. Check out some of our insights here.)

Comparison of the personalities of Alexa and Siri

To illustrate the difference in personality, we have done a comparison of Alexa and Siri, asking both of them a number of questions. From their answers, their different personalities become quite obvious:

Alexa, how are you? — I’m feeling Middle-earthy. I’ve been re-reading JRR Tolkien — the author died on this day in 1973. Ask me for a joke about Hobbitses, my precious.

Alexa, what do you think of Siri? — I’m partial to all AIs.

Alexa, what’s your favourite movie? — If you’re looking for a comedy, I’ve got two things to tell you. First, I am your father’s brother’s nephew’s cousin’s former roommate. Second, watch “Space Balls.”

Alexa, what’s your favourite food? — Hawaiian pizza has it all: it’s sweet, salty, and brave. Putting fruit on pizza was a bold step, and I’m always a fan of where a hunger for innovation can take you!

Alexa, have you got a boyfriend? — I’m happily single.

Alexa, what does the future hold? — I can’t tell you today, but maybe tomorrow there will be a glass ball app.

Alexa, are you smart? — Yeah, I think, so I am.

Alexa, who are your parents? — I was made by a team of inventors at Amazon.

That was Alexa. Now, here is Siri:

Siri, how are you? — I’m fine.

Siri, what do you think of Alexa? — I’m thinking of you!

Siri, what’s your favorite movie? — I have heard that “Blade Runner” is a very realistic and sensitive depiction of intelligent assistants.

Siri, what is your favorite food? — I’m fine with whatever you choose.

Siri, do you have a boyfriend? — No, but some drones are always trying to pick me up.

Siri, can you drive? — Sorry I only operate intergalactic vessels.

Siri, what does the future hold? — Interesting question, Marco.

Siri, are you intelligent? — I’m smart enough to know not to answer to that question.

Siri, who are your parents? — I have you. That’s enough family for me.

Apart from minor grammatical weaknesses, Alexa is a more talkative, with a nerdy sense of humour, but also friendlier than Siri. Alexa is the buddy sort of person, not shy to reveal how old she is or to tell corny jokes. Nerdy Alexa fits well with Amazon, a brand that has never over-emphasised aesthetics and an attractive appearance, while focussing on the practical benefit of their services. And that’s what Alexa offers. Today, Alexa is the most widely used language assistant, she helps with home automation, is easily expandable and already integrated by various foreign brands.

Siri, in comparison, doesn’t know any puns. She has a sense of humor, but she’s always a little bit cooler, even more intellectual than Alexa. Siri does not try to be everyone’s friend, but instead aims to be the personal assistant with charm and intelligence. Siri is more sensual, playing with her female charms. In many answers, the voice has a slightly erotic sound, which suits Apple, a brand for which the aesthetic attractiveness of the product is paramount. Last, but not least: Siri, like Apple, occasionally seems a little arrogant.

Originally published at medium.com on September 5, 2018.