Artificial Intelligence

What Is a Branded AI Assistant?

Posted on 23 March 2026 by think moto

AI assistants are becoming a central interface between companies and customers.
This shift is changing how brands need to be designed.

For a long time, brands were primarily created for communication: websites, campaigns, and visual identities. Today, brands are increasingly beginning to interact. AI assistants answer questions, explain products, and help people make decisions. As a result, something fundamental is changing: the brand begins to speak. Once a brand communicates through AI, it is no longer just a piece of technology.
It becomes part of the brand experience.

This development marks a turning point in brand management. While digital transformation previously meant optimizing brands visually for screens, the new challenge is to design brands for conversation. This is not a gradual extension of existing design practices. It is a categorical shift: from representation to interaction, from presentation to dialogue.

The Shift in Digital Brand Management

Over the past decade, digital products have already transformed how brands operate. Websites became platforms. Products became services. Interfaces became the central place where brand experience happens.

Now, AI assistants introduce a new type of interface that differs fundamentally from previous ones: they communicate not only visually, but linguistically. With large language models, companies can develop assistants that answer complex questions, guide users through services, or support decision-making. What previously required menus, forms, or support hotlines can now happen through natural language conversations. An insurance customer, for example, no longer needs to navigate a complex form. Instead, they simply describe their situation. The assistant understands, asks follow-up questions, and explains available options.

Conversational interfaces do something that traditional digital interfaces rarely achieved: they communicate in natural language, with all the nuances that come with it — tone, attitude, and personality. This fundamentally changes the role of brands. Because as soon as an organization participates in conversations, it reveals how it thinks, argues, and explains. The brand is no longer just seen.
It is experienced — in every answer, every clarification, and every explanation.

The Blind Spots in Current AI Implementations

Many organizations currently see AI assistants mainly as technical tools to improve efficiency: automating support, reducing costs, or answering frequently asked questions. From an operational perspective, this makes sense. From a brand perspective, however, something critical is often overlooked.

An AI assistant represents the company. It explains products, responds to criticism, and helps users navigate complex decisions. In many cases, it speaks on behalf of the brand more frequently and more directly than any marketing campaign ever could. If this voice is not deliberately designed, inconsistencies quickly emerge that fragment the brand identity.

Different assistants speak differently. Responses vary depending on prompt engineering. Brand positioning becomes diluted because different teams maintain different knowledge bases. In visual brand management, consistency is standard practice: typography, color systems, and imagery are carefully defined. In conversational interfaces, this level of discipline is often missing. As a result, brands that spent years building a coherent visual identity suddenly speak with ten different voices.

Branded AI Assistants: A Conceptual Framework

This is where the concept of the Branded AI Assistant comes in. A Branded AI Assistant is more than a chatbot connected to a knowledge base. It is a deliberately designed interaction layer between an organization and its users. Several dimensions shape this layer:

Brand Voice: The assistant does not simply provide correct answers. It communicates in the characteristic tone of the brand. If a brand is precise and factual, the assistant responds with clear, structured explanations. If it is approachable and encouraging, it explains patiently, asks clarifying questions, and provides helpful context.

Conversational UX: Dialogues are systematically designed rather than left to chance. This means anticipating conversation flows, identifying common user intentions, and developing consistent response patterns.

Personality: The assistant has a defined way of reacting. How does it deal with uncertainty? How does it admit mistakes? How proactively does it guide the user? This personality is not a property of the AI model. It is a design decision.

Governance: Knowledge sources must be curated. Responses need to be reviewed regularly. Prompts should be maintained systematically. This requires clear responsibilities and processes — similar to content governance in traditional digital ecosystems.

Interaction Principles: Rules define how the assistant explains, guides, and responds. Does it answer immediately or ask clarifying questions first? How much context does it provide? How direct are its recommendations?

Only when these dimensions are consciously designed does a technical solution become a true brand interface. The difference is comparable to the one between a functional website and a carefully crafted digital brand experience.

Practical Implications for Organizations

As AI assistants become brand interfaces, responsibilities inside organizations begin to shift. Brand teams, design teams, and product teams need to collaborate more closely than before. Questions that used to be either technical or creative now become both:

How does the brand explain complex topics? How does it respond to criticism or complaints? How actively does it guide users through decisions? How does it handle uncertainty? These questions shape the brand experience just as strongly as typography, color systems, or visual language. Therefore, they increasingly belong inside brand systems, not only in technical architectures or prompt libraries.

In practical terms, companies must define their AI voice as systematically as their visual identity. They need to establish conversational design as a discipline. And they must create governance structures that ensure consistent AI interactions across touchpoints.

First Steps for Companies

Organizations that want to design AI assistants strategically can begin with several concrete steps. First, define the AI voice. This translates the brand’s tone into conversational rules. It does not mean simply copying existing brand guidelines, but clarifying how the brand sounds in direct dialogue. How much personality does it express? How formal or accessible is it?

Second, establish conversational design as its own discipline. This includes designing dialogue flows, defining typical conversation patterns, and developing interaction principles. Unlike traditional user interfaces, the focus here is not on click paths but on conversation dynamics — including the uncertainty and variability that natural language brings.

Equally important is the establishment of clear governance structures. Responsibilities for content, prompts, and knowledge sources must be defined. Processes for regular review and optimization should be implemented. Finally, AI interactions should be integrated into existing brand systems, alongside design systems, brand guidelines, and product design frameworks.

Only through this structured approach does a technical tool become a consistent part of the brand — an interface that not only works, but strengthens the brand identity instead of fragmenting it.

A New Design Challenge for Brands

For a long time, brands were primarily designed for visibility — to attract attention, create recognition, and establish visual differentiation. In the age of AI, brands are increasingly designed for interaction. This is more than a technological development. It represents a fundamental expansion of what brand management means.

Visual identity defines how a brand looks. Conversational identity defines how it thinks, argues, and communicates. It reveals how an organization understands problems, structures decisions, and deals with complexity. In this sense, Branded AI Assistants are not just a new technology interface. They are a new medium of brand management.

The challenge for organizations is not to leave this new dimension to chance, but to design it as deliberately as every other aspect of their brand. Not only defining how a brand looks — but how it speaks.

Customer Journey in the Age of AI: When Brand, Touchpoints and Technology Converge

Posted on 17 February 202625 February 2026 by think moto

Marken existieren nicht isoliert. Sie entstehen im Erlebnis des Kunden und verändern sich permanent. Die Markenpersönlichkeit sollte an jedem Kontaktpunkt erfahrbar sein und im Zusammenspiel ein stimmiges Bild ergeben. Dabei geht es weniger um starre Konsistenz als um kohärentes, dem jeweiligen Kontext angepasstes Verhalten. Und genau hier verändert künstliche Intelligenz das Spiel fundamental.

Die Customer Journey – die Reise vom ersten Markenkontakt bis zur langfristigen Kundenbindung – war schon immer komplex. Doch mit KI-gestützter Personalisierung, intelligenten Assistenten und datengetriebenen Entscheidungen in Echtzeit erreicht diese Komplexität eine neue Dimension. Marken, die heute erfolgreich sind, verstehen, dass ihre Touchpoints nicht mehr statische Kontaktpunkte sind, sondern dynamische Interaktionsmomente, die sich dem individuellen Nutzer anpassen.

Von der Analyse zur intelligenten Gestaltung

Customer Journey Optimierung beginnt mit der ehrlichen Bestandsaufnahme: Wie erleben Kunden unsere Marke tatsächlich? Wo entstehen Reibungsverluste? Welche Momente führen zu Freude, welche zu Frustration? Diese Current State Journey zu verstehen ist der erste Schritt. Der zweite ist mutiger: Wie könnte die ideale Reise aussehen? Welche Touchpoints wollen wir schaffen, die heute noch nicht existieren?

Das Ziel ist klar – Reibungsverluste eliminieren, die User Experience personalisieren, Conversion Rates steigern. Doch das Wie hat sich verändert. Früher optimierten Marken ihre Touchpoints basierend auf Durchschnittswerten und A/B-Tests. Heute ermöglichen KI-Systeme die Personalisierung für jeden einzelnen Nutzer in Echtzeit. Das ist nicht nur eine technische Evolution, sondern eine grundlegend neue Art, Marke zu denken.

Die fünf Phasen der Journey – neu gedacht

Die klassische Einteilung in Awareness, Consideration, Decision, Post-Purchase und Loyalty bleibt relevant. Aber innerhalb jeder Phase verschiebt sich der Fokus. In der Awareness-Phase geht es längst nicht mehr nur um Sichtbarkeit durch SEO und Content Marketing. Predictive Analytics können heute potenzielle Kunden identifizieren, bevor diese überhaupt aktiv suchen. KI-gestützte Content-Generierung ermöglicht skalierbare Personalisierung für verschiedene Zielgruppen, ohne dabei an Qualität zu verlieren.

In der Consideration-Phase, in der Kunden verschiedene Optionen vergleichen, spielen nicht mehr nur Produktbewertungen und Testimonials die entscheidende Rolle. Intelligente Recommendation Engines schlagen passende Lösungen vor, oft bevor der Kunde selbst weiß, wonach er sucht. Interaktive Konfiguratoren und KI-Chatbots begleiten die Evaluierung und schaffen Vertrauen durch unmittelbare, präzise Antworten.

Die Entscheidungsphase wird durch technische Exzellenz bestimmt – optimale Ladegeschwindigkeit, intuitive Navigation, klare Call-to-Actions. Doch auch hier arbeitet KI im Hintergrund: Dynamische Preisanpassungen, personalisierte Angebote und Predictive Lead Scoring erhöhen die Wahrscheinlichkeit einer Conversion, ohne dabei manipulativ zu wirken. Es geht um den richtigen Impuls im richtigen Moment.

Nach dem Kauf beginnt für viele Marken die eigentliche Arbeit. Hervorragender Service, personalisierte Kommunikation und proaktive Problemlösung liefern die Basis dafür, dass Kunden wiederkehren. KI-gestützte Support-Systeme ermöglichen 24/7-Verfügbarkeit, während Sentiment-Analyse Unzufriedenheit frühzeitig erkennt und menschliche Customer Service Agents gezielt einschaltet. Automatisierte Call-Zusammenfassungen entlasten Support-Teams und ermöglichen es ihnen, sich auf komplexe Fälle zu konzentrieren.

Die Loyalty-Phase ist die wertvollste. Hier werden Kunden zu Markenbotschaftern. Intelligente Treueprogramme, die auf individuellem Verhalten basieren, Community-Building und Predictive Retention sorgen dafür, dass diese Beziehung nicht abbricht. Denn loyale Kunden sind wertvoller als neue – Retention ist günstiger als Akquise.

Customer Journey Maps als lernende Experience-Systeme

Customer Journey Maps sind ein wirkungsvoller Ausgangspunkt für die kontinuierliche und agile Verbesserung von Kundenerlebnissen. Sie zeigen nicht nur, wo Menschen mit einer Marke interagieren, sondern auch warum, wie und mit welchen emotionalen Reaktionen. So werden Moments of Truth sichtbar – jene kritischen Situationen, in denen sich Wahrnehmung, Vertrauen und Erfolg entscheiden.

Eine professionelle Journey Map erfasst Touchpoints über den gesamten Customer Life Cycle, beschreibt User Tasks und den jeweiligen Nutzungskontext. Sie macht Brüche, Reibungen und Pain Points ebenso erkennbar wie funktionierende Interaktionen und echte Mehrwerte. Vor allem aber zeigt sie, ob die Markenpersönlichkeit entlang der gesamten Journey konsistent erlebbar wird.

Doch Journey Maps sollten nicht beim Status quo stehen bleiben. Als lernende Experience-Systeme helfen sie, Potenziale zu modellieren, Hypothesen zu testen und Interaktionen gezielt weiterzuentwickeln. Gutes Design kann Verhalten verändern – vorausgesetzt, Marken haben den Mut, neue Wege auszuprobieren. Denn Kund:innen können nur bewerten, was bereits existiert. Zukunftsorientierte Marken gestalten nicht nur bessere Journeys – sie schaffen völlig neue Erlebnisse.

KI als Katalysator, nicht als Ersatz

Künstliche Intelligenz verändert die Customer Journey fundamental. Aber nicht, indem sie Menschen ersetzt, sondern indem sie menschliche Entscheidungen augmentiert. KI-Systeme können Verhalten vorhersagen, in Echtzeit personalisieren und Support automatisieren. Doch die strategische Gestaltung der Journey, die Definition von Markenwerten und die Kreation neuer Touchpoints bleiben menschliche Disziplinen.

Intelligente Systemlösungen wie Recommendation Engines, Personalisierungstools oder Conversational AI sind Werkzeuge. Sie ermöglichen es Marken, datengetrieben und agil zu optimieren. Aber sie ersetzen nicht das strategische Denken, das hinter einer kohärenten Markenerfahrung steht.

Human-First. AI-Backed – das ist der Ansatz, der funktioniert.

Der Weg ist das Ziel

Customer Journey Optimierung ist kein Projekt mit Enddatum. Es ist ein kontinuierlicher Prozess. Marken, die systematisch Daten analysieren, auf Kundenbedürfnisse eingehen und gleichzeitig den Mut haben, neue Wege zu gehen, schaffen nachhaltige Wettbewerbsvorteile. Die Kombination aus strategischem Journey Mapping, datengetriebener KI-Optimierung und mutigem Design schafft Erlebnisse, die nicht nur konvertieren, sondern begeistern.

Eine Marke ist keine Insel. Aber die Reise entlang ihrer Kontaktpunkte sollte einzigartig sein.

Möchten Sie mehr darüber erfahren? Sprechen Sie mit uns über Ihre Herausforderungen – wir verbinden strategisches Brand Thinking mit modernster KI-Expertise.

Mehr zu unserer Expertise in Brand Intelligence

Why Conversational AI Is Now Part of Corporate Design

Posted on 27 January 202629 January 2026 by think moto

Sprachassistenten, Conversational Interfaces, Agentic AI, Customer Service Automation, In-Car Assistants, Mobile Apps, Websites, ja sogar Produkte und Geräte selbst bieten die Möglichkeit zur sprachlichen Interaktion.

Wenn diese Interfaces nicht zur Markenidentität passen, entsteht Reibung: Ein visuell hochwertiges Corporate Design trifft auf einen generischen Bot, der spricht wie jede x-beliebige KI? Das zerstört Vertrauen — und wirkt unprofessionell.

Conversational Design löst dieses Problem, indem es:

Tonalität, Wortwahl und Satzrhythmen der Marke definiert
die Persönlichkeit der Marke in eine dialogische Form bringt
die Brand Voice in Chatbot- und Voice-Umgebungen konsistent macht
Kundenerlebnisse emotional stimmiger, klarer und intuitiver gestaltet
Teil der Integrated Brand Experience wird

Kurz gesagt: Conversational Design übersetzt Corporate Identity (CI) und Corporate Design (CD) in Branded Conversations.

Von der visuellen zu einer dialogischen Corporate Identity

Eine zeitgemäße Markenidentität setzt sich heute aus drei miteinander verbundenen Ebenen zusammen: der visuellen Ebene, der sprachlich-narrativen Ebene und der dialogischen Identität. Logos, Typografie und Farbwelten definieren weiterhin das Erscheinungsbild. Purpose, Messaging und Tonalität bestimmen, wie eine Marke klingt.

Die dialogische Identität schließlich übersetzt all das in konkrete Interaktionen: Wie formuliert der digitale Assistent Antworten? Welche Haltung spricht aus kurzen, funktionalen Sätzen? Wie wird Missverständnissen begegnet? Welche wiederkehrenden Muster, Mikro-Formulierungen und Gesprächsprinzipien prägen den Kontakt?

Conversational AI schafft damit den Übergang von einer statischen Markenwelt zu einem lebendigen, interaktiven System. Es sorgt dafür, dass jede Konversation — ob im Chat, per Voice oder in hybriden Interfaces — konsistent mit der Markenpersönlichkeit bleibt. So wird die Identität nicht nur gesehen und gelesen, sondern auch erlebt.

Praxisbeispiele

Für HUGO BOSS haben wir die dialogische Ebene des Style-Assistants entwickelt: ein Assistant, der Modekompetenz, Selbstbewusstsein und die elegante Direktheit der Marke in einer klaren, markenspezifischen Tonalität abbildet.

Auch bei Audi entstand ein Conversational UX Framework, das die ruhige Präzision und technische Klarheit der Marke in die Sprache ihrer Voice- und Assistenzsysteme überträgt.

Und für ein führendes deutsches Unternehmen im Bereich Prüfung und Zertifizierung haben wir kürzlich ein komplettes Webinterface dialogisch gestaltet — jeder Interaktionsschritt beginnt dort mit einer markentypisch formulierten Konversation, statt mit klassischen UI-Bausteinen. Das Ergebnis: ein Nutzererlebnis, das intuitiver wirkt und die Marke in jeder Interaktion spürbar macht.

Wie man Conversational Design systematisch verankert

Bei think moto integrieren wir Conversational Design heute standardmäßig in Markenprozesse. Der Workflow umfasst:

1. Conversational Identity Definition

Übertragung der Markenpersönlichkeit auf die AI
Sprache, Tonalität, Satzstrukturen
Do’s & Don’ts
Response Patterns

2. Dialog-Module & UX Patterns

Intent-Strukturen
Interaktionsmodelle
Micro-Conversations
Error Handling

3. Technische Übersetzung

Prompting-Guidelines
Trainingsdaten
Knowledge Architectures
Integration in LLM-, Voice- oder Chatbot-Systeme

4. Brand AI Governance

Conversational Styleguide
Scalable Prompt Framework
Cross-Touchpoint Consistency

Fazit: Eine Marke muss heute sprechen — und zwar in ihrer eigenen Stimme

Conversational Interfaces werden in den kommenden Jahren einer der wichtigsten Berührungspunkte zwischen Marken und Menschen sein. Wer dort generisch erscheint, verliert. Wer dort markentypisch, empathisch und konsistent spricht, gewinnt Vertrauen, Nähe und Relevanz. Conversational Design ist deshalb kein technisches Thema, sondern ein Marken- und Corporate-Design-Thema.

Human First. AI-Backed.

Das gilt besonders für Marken, die in einer KI-geprägten Zukunft bestehen wollen.

Erfahre mehr über unsere Arbeit mit AI für Marken unter thinkmoto.de/KI-Branding und unter thinkmoto.de/Chatbots.

Why branding for the industrial Mittelstand is more critical than ever

Posted on 22 January 202627 January 2026 by think moto

Germany’s Mittelstand is widely seen as the backbone of the economy: highly specialized, technology-driven, and globally competitive through exports. Yet while machinery, materials, and production lines are continuously upgraded, one area often falls behind: the brand.

Many mid-sized industrial companies invest in branding, corporate design, or brand strategy only sporadically – typically when a relaunch is due or competitive pressure intensifies. In between, things often stand still. But this standstill is costly.

How the Mittelstand manages branding today—and why it’s becoming a problem

In many industrial companies, brand management still follows a traditional model: external agencies develop corporate designs, create guidelines, review campaigns, and run competitor analyses or brand audits. Internally, small marketing teams handle day-to-day execution and try to keep long-term brand development on track.

These structures have grown over time – but they come with three fundamental weaknesses:

1. Project-based, not continuous.
A corporate design gets updated – yet no routine follows to maintain it consistently over years.

2. High costs, limited scalability.
Every analysis, every adjustment, every approval requires new external budgets, time, and coordination.

3. Insufficient use of strategic brand work.
Because agency services feel costly, leadership often decides against them – and accepts the gradual erosion of the brand.

The result is visible across many industrial sectors: inconsistently designed channels, divergent layouts, fragmented brand messages, and products that feel more interchangeable than they actually are.

Interchangeability is the biggest risk for the industrial mittelstand

The frequently cited McKinsey analysis “Late vs. Made in Germany” highlights the following conclusion:a lack of brand leadership leads to commoditization. When products and services are technically world-class but not clearly differentiated visually, verbally, or strategically, Mittelstand companies compete almost exclusively on price and functionality.

This is strategically risky, because commoditization leads to:

increasing price sensitivity
declining customer loyalty
higher marketing and sales costs

And yet Mittelstand industrial companies would be perfectly positioned to build strong brands in line with our concept of Spherical Branding: Deep expertise, technological excellence, quality, mindset, and values form an ideal foundation for credible differentiation.

The real cost: high effort vs. high loss

Direct costs:

Recurring agency fees for layout checks, design adaptations, and brand reviews
Unclear processes that lead to long approval cycles
Small marketing teams drowning in operational workload

Indirect costs (often bigger):

Blurry brand presence across different marketsoutdated messages that no longer fit the company’s strategy
Inconsistent presentations, websites, and product communication
Long-term brand weakening and declining perceived quality
Increasing need for expensive relaunches

Why branding is more important for industrial companies than ever before

Digital transformation, new competitors from Asia, skilled labor shortages, and global pricing pressure are changing the rules.

Brands that are clear, consistent, and differentiated benefit in several ways:

Stronger competitive positioning
Higher visibility across digital channels
Clearer value propositions
Greater employer branding
Stronger pricing power
Closer customer relationships – including AI-based touchpoints

In a world where data, interfaces, and machines increasingly shape interactions, the brand must remain recognizable as the human layer: empathetic, credible, and distinct.

Rethinking brand management: Human First. AI-Backed.

Modern brand leadership in the industrial Mittelstand requires two things:

1. Strategic clarity and identity.
a brand must know who it is – what it promises, how it speaks, and what it looks like.

2. Support from intelligent systems.
The future of branding is hybrid: human creativity + AI-powered tools that make processes more efficient, reveal data patterns, accelerate workflows, and secure brand consistency.

This makes branding not only more emotional, but also more precise, scalable, and economically viable for mid-sized companies.

Conclusion: the mittelstand doesn’t need more branding – it needs better branding

Branding is not a nice-to-have. It’s a strategic value driver that determines whether companies remain visible, relevant, and differentiated in the future.

The good news: it has never been easier to build a lean, data-driven, and future-ready brand system than it is today – Human First. AI-Backed.

And this is exactly where a major opportunity begins for the industrial Mittelstand.

Human First. AI-backed.

Posted on 5 January 2026 by think moto

Why Brands Are Becoming Human Again. The last few years belonged to technology. The years ahead belong to people — precisely because technology has become so powerful.

With Human First. AI-backed., we articulate our stance for a future in which AI does not replace humans, but amplifies them. It does not dominate — it empowers.

Human First stands for responsibility. And for radical creativity.

Brands must reconnect with emotion, learn to listen again, and create real meaning.
It’s about empathy, user-centered thinking, and the courage to make clear decisions. Intuition. Imagination. Judgment. These remain fundamentally human.

AI-backed means we design differently — and we advise differently.

We have rethought every step of our workflow: research, strategy, naming, brand voice, design, prototyping. AI changes speed and quality. We have rethought every step of our workflow: research, strategy, naming, brand voice, design, and prototyping. AI changes speed — and it changes quality.

Brands today are built within integrated, intelligent design systems. Strategies become sharper. Brand experiences more adaptive. Agentic AI solutions open up entirely new dimensions of brand leadership.

One thing is becoming unmistakably clear: it’s not the size of a team that matters, but the seniority of the minds behind it. AI amplifies what already exists. It does not replace responsibility.

That’s why we invest in experience, depth, and creative excellence — supported by purpose-built intelligent systems. This is how brands become not just more consistent, but more alive. The future of brand leadership is not about choosing between human creativity and technology. It lies in their interplay.

Human First. AI-backed.

Connecting with Microsoft AI and NVIDIA in Berlin

Posted on 20 March 202430 November 2024 by Paul Krizsan

When Microsoft announced that they would be in the neighborhood with one of our favorite topics, we knew we had to attend the Microsoft AI Tour stop in Berlin. With an agenda full of fascinating topics it was hard to choose which sessions to attend for our technology and AI team.

The tenor of the event has been one of reliability and maturity. Microsoft and their event partner NVIDIA wanted to let everyone know that artificial intelligence is long past the experimental stage of proof-of-concepts and haphazardly assembled tech demos. AI has reached a point wherein scalability and establishment of best practices and operations is now more critical than sheer feasibility.

In this post we have broken down four of the key areas and what we have observed and learned.

Strong commitments to security and trust

“How many of you trust AI?” started the speaker, and the answer became evident upon counting the hands in the audience–not many. Most AI users usually don’t really know what is happening under the hood, and it is difficult to predict the output. Even the researchers and scientists behind the technology are constantly discovering new behaviors and peculiarities of their models, not seen before.

As we all know, LLMs are usually trained on data from across the internet, thus prompting the clear question: What do they know about me, and if they will use what I said when talking to other users?

At this point, we all have many questions about our privacy when using AI. Therefore, Microsoft has unveiled the curtain behind their approach to securing AI and has shared what should be taken into account when building an AI-powered application.

At their core, LLMs and foundation models that have not been fine-tuned are difficult to trust: unpredictable hallucinations and with few guard rails to reign in their output.

However, using additional layers of safety on top of the foundation models, such as grounding it with high-quality knowledge bases and self-evaluating patterns and prompt instructions, developers can define safe corridors for reliable and consistent output and responses.

Safety is a critical component that stretches across the entire bandwidth of generative AI: from the data ingested to instructions during processing of input and output, all the way to disclaimers and labels in the user interface.

So our question should be reframed: Do I trust the developers behind the specific AI tool?

Apart from discussing general concepts and best practices, Microsoft has introduced its new Security Copilot, which will be available from May 1st and can function as a SOC Analyst. It can answer questions about the state of your Microsoft apps connected to the cloud, conduct security audits, and assist in debugging various application errors. Moreover, it can even alert you if an email you receive contains an attempted phishing attack.

Copilots, not autopilots

The vibrantly enthusiastic Seth Juarez summed it up perfectly: the current suite of AI that Microsoft can provide is not meant to run all on its own. You need to assist it in refining the desired output, guide it and tell it what worked and what maybe is not correct.

With multiple breakout and workshop sessions on how to build your own copilot throughout the day being completely overrun and people being turned away from overflowing rooms, it was obvious what topic is on everyone’s mind.

From copilots that monitor your systems and logs to report on potential vulnerabilities and breaches to copilots that summarize analyze business data and create reports with suggestions. The possibilities are endless and nearly every Microsoft product is now equipped with it’s own copilot feature.

But what is clear is that this plethora of different entry points into conversational interfaces also seems convoluted and raw. UX patterns and output formats vary greatly and the question arises: wouldn’t it be a lot more accessible and friendly to have a singular interface with the AI rather than having to jump between them constantly?

A better solution will likely be OS-integrated LLMs that provide standardized patterns or modules that application can tap into, but we are not quite there yet. Until then, every service and product will likely have its own standalone conversational interface before they will be consolidated in a swift move on the OS level.

RAG and AI search

Vector search and retrieval-augmented generation (RAG) still has generative AI firmly in its grasp. The capabilities of vectorized databases for grounding are profusely necessary for reliable output and a significant reduction of hallucinations.

Both Microsoft and NVIDIA presented models and best practices for workflows for preprocessing and referencing of vector databases as well as demonstrations for Azure AI studio.

While it seems as though technological advancements have slowed a bit in this field recently, there were some takeaways for us that illustrated the necessity for indexing when handling large amounts of documents for the embedding model.

Establishing LLMOps for production-ready applications

What fell short in my eyes was the sharing of best practices on how to enable less tech-savvy people in organizations and provide building blocks for prompts or the development of pipelines of building blocks for prompts.

One of the biggest challenges of establishing acceptance of LLMs and AI within organizations is the fact that, while we are communicating on an intuitive level, natural language interfaces are not something we are accustomed to. Learning how to properly phrase prompts with their intransparent intricacies of what you should include and how, should be managed and accompanied with tooling that steers users accordingly.

A substantial part of building the right infrastructure should be the automated inclusion–or at least privision–of repeatedly and frequently used excerpts and prompt modules that get attached to prompts. If I am generating content for my marketing materials, I should not constantly have to paste instructions for the tone of voice of the output or camera settings for consistent image generation.

While Microsoft covered potential workflows for rewriting user queries before generating the final output, a lot of struggles and challenges of these enterprise-level ops management and tasks was unfortunately not mentioned.

In conclusion

Overall we greatly enjoyed the event, meeting fascinating people (some with the obligatory and adorned Apple Vision Pro) and listening to what drives and motivates others in the realm of artificial intelligence these days.

Thank you Microsoft and your efforts in outreach to developers and advocates!

Reflections on the world’s first Conversation Design Conference UNPARSED

Posted on 31 August 202330 November 2024 by Susann Pocha

The debut of the Unparsed Conference this year was a testament to the dynamic evolution of the field of Conversation Design. With the advent of AI-driven language models, the landscape has transformed dramatically, raising intriguing questions about the nature of our profession and its relationship with the capabilities of Large Language Models (LLMs).

Evolution of Conversation Design: From Decision Trees to Neural Networks

The conference prominently highlighted the journey of Conversation Design, tracing it from its humble beginnings in decision trees to its current manifestation in intricate neural networks. This evolution demonstrated the remarkable progress our profession has made and set the tone for exploring new avenues and frontiers in conversation design. The most compelling questions were: What constitutes conversation design today? How do our human skills differ from the skills of LLMs? How will Conversation Experts work in the future? What will be their value-add?

This question also drives us at think moto. We see a big shift from crafting conversations to engineering prompts. We are curious to explore all possibilities to develop prompt architectures that can perform any conversation with any specific content.

A Glimpse of the Possibilities of the Future

UNPARSED discussed innovative prototypes, from AI-generated word games to chatbots that can engage in complex philosophical debates. The 20-minute lectures encouraged thinking about AI companions that could truly engage in conversations and raised the exciting prospect of conversational interactions that go beyond mere assistance. Our key takeaway here: When designing these assistants, it is no longer the variations of possible user questions that are the training data, it is the content itself. The better the content provided to the LLMs, the better, more authentic, and more appropriate the answers.

Ethical Considerations

Panel discussions addressed the ethical dimensions of conversational design, reflecting the increasing intertwining of AI and human interaction. The debate over whether AI discloses in advance what sources are used or how trustworthy a statement is being made is deep and mature. There is a consensus that, on the one hand, human education is crucial, and on the other hand, a responsible approach to the use of technology.

Conversation Design as an Art

The heart of conversational design lies not in mere lines of code or algorithms, but in the art of designing meaningful interactions. The user is always at the center. The conference reinforced that technology should adapt to human behavior, reaffirming that conversation design is an art that requires a deep understanding of human nuances and psychology. At think moto, we believe that the process of conversation design is changing, but the result should be useful, helpful, and enriching, as always. This still requires good research, strategy, and an appropriate approach for each use case.

The UNPARSED 2023 conference was an exciting journey that broadened horizons and demonstrated the potential of conversational design. The diverse inputs from the lectures once again demonstrated the importance of lively debate – whether human, artificial, or a fascinating mix of both – continues to shape the way we interact and communicate.

And what influence we as conversation designers should have on what’s technically possible. Because those who use the technology should question it.

We are excited to see how the field of conversation design will continue to evolve and are looking forward to UNPARSED 2024.

Why Content Designers and AI Could Actually Be a Dream Team

Posted on 27 July 202330 November 2024 by Susann Pocha

If you’re a content designer, UX writer, or hold any other role in the field of the ever-evolving world of digital content creation, the prospect of integrating artificial intelligence (AI) into your work process is both exciting and daunting.

Large language models (LLMs), in particular, are passionately debated for how they will not only transform our work but completely disrupt it. The possibilities for editing and generating text are diverse and evolve rapidly. Yet, in the future of content design, human creativity, and AI efficiency are not competitors but rather dynamic teammates. Here are three reasons why content designers should seriously consider pairing up with AI.

1. AI is efficient and effective

AI can significantly improve the efficiency and effectiveness of content design. But only if content creators understand how AI works. The better they understand the functionality and capabilities of the system, the better they will generate useful prompts that transform their ideas in real-time. Also, AI systems are outstanding at mining vast quantities of data and identifying patterns. This can save time or open up completely new perspectives. The once-dreaded blank page can lose its intimidation. Knowing these AI capabilities and how to use them paves the way for a more efficient and effective content-creation process.

2. Humans can decide the focus

AI can free content designers to concentrate on what humans do best: shaping ideas, strategizing, and solving problems with empathy and purpose.

While AI can manage the routine tasks of data analysis, pattern recognition, and interpreting insights, it’s humans who lead the way. It’s humans who provide the perfect briefing for the artificial teammate, ensuring it creates content that is not only accurate but also relevant and ethical. It’s perfectly clear: Crafting a well-thought-out strategy is essential to nourish the AI with precisely tailored prompts and get it reflect on the ideas you’ve been thinking about. This thoughtful application of AI could help generate more targeted and personalized content, resulting in better user engagement. The more focused the thinking of humans, the more valuable the result of the AI.

3. Writing skills become a new benchmark

While AI can create mostly generic content, it still can’t replicate the exceptional writing skills that humans possess–at least not without inhuman efforts to engineer prompts.AI lacks the ability to fully comprehend nuances of language, emotion, and culture that are often essential for compelling content.

AI fails to adapt content to user and business needs. This is where content designers shine, by giving communication a fresh touch that makes it exceptional, if not surprising!

So … what’s next?

To wrap it up: AI’s role is not to replace content designers, but rather to augment their capabilities. For me, it offers a way to automate repetitive tasks so I can focus on strategic thinking, problem-solving, leadership, and maintaining exceptional writing quality. When designing conversational experiences for our customers, my role has shifted. I no longer point out the many constraints within the conversation that the chatbot used to have, but instead, give constraints to the AI to keep the conversation on track and aligned with user and business needs.

It’s incredibly fun and challenging at the same time to put yourself not only in the shoes of the user but also in the shoes of AI. As this process is just emerging, it is up to us content designers to use this service wisely and thus shape the future of our profession.

What ChatGPT and LLMs Mean for How We Build Conversational Interfaces for the Future

Posted on 28 February 202330 November 2024 by Paul Krizsan

Firstly, what are LLMs and ChatGPT? This is not an article about what LLMs (or large language models) and ChatGPT are. If you have been living under a rock and are unfamiliar with these names and terminology then this article written by ChatGPT explaining itself should be a good starting point.

We have been receiving questions from – and participated in many discussions with – our customers and peers about this exciting new tech and wanted to clarify our stance on where we see the opportunities and weaknesses at the current stage, as well as looking forward to a potential hybridized future. The biggest talking point has been the need for conversation design in an increasingly automated and generative world.

From our perspective as experts on conversational interfaces and conversation design we see predominantly two paths that this technology and trend will continue to develop on: the path of consumer-facing applications and the path of the technology as a tool and force multiplier. Neither of which will be eliminating the need for humans behind the wheel, steering the technology, anytime soon.

Hopping on the LLM bandwagon

Broadly speaking, this technology and its implications are spreading at breakneck speed. Many platforms are currently aiming at capitalizing on this goldrush-like state. You may have heard of Microsoft implementing ChatGPT in Bing and Google looking at fusing their proprietary equivalent LaMDa with their own search engine. These search engines follow a trend that companies such as SoundHound have been pursuing for a while, responding to users not in lists of search results, but in concrete answers in the form of natural language.

Other examples of quick wins in this brand new space are bot platforms such as Voiceflow and Cognigy.AI. Here the same purpose of applying LLMs to dynamically generate the system responses or predictable training data for intent training is being used heavily. Some platforms, like Cognigy.AI, are also considering going a step further and looking into the empowerment of conversation designers by allowing the creation of flows and elements through natural language prompts, speeding up the process of setting up new conversations greatly and thus contributing to rapid prototyping capabilities of these low-code platforms. Will these features collate into conversations that are production-ready, about to be rolled out to millions of users, out-of-the-box? Of course not. But they provide a good first framework to expand upon.

Trust in the system and the tech is dwindling

Widely broadcasted anecdotes of tech journalists and influencers, as well as hear-say from colleagues and friends have recently lead to a lot of skepticism when it comes to the current state of the technology. Articles quoting the unsettling feeling, individual erroneous responses and behavioral patterns reinforce negative connotations when it comes to LLMs in todays world. This obviously has a huge negative impact on consumer-facing applications.

Finding an appropriate place for LLMs should not be difficult

Focusing on this new technology as a force multiplies and enablement tool, is therefore the more stable path from our perspective. At least while the technology matures and new, more refreshing experiences for consumer-facing applications improve the publics perception in the mid-term.

On a more immediate and applied note, ChatGPT and LLMs are a great vehicle for innovation and a popular driver for change, but they are tools and will not replace human experts in conversation design. It is a good gap-filler and repetitive tasks but it will not provide the confidence and accuracy of dialogues designed by humans for a while.

The conversation designer is still the agent of change for this new tech

Our workflows in the future could consist of conversation designers laying down the structure of a dialogue, such as the starting point, the goal of the conversation and some checkpoints along the way, with the generative AI or LLM filling the gaps.

In an ideal world we would provide the AI with a purpose and a personality, but no actual dialogue would need to be written by humans. The conversation designer would be focused entirely on the strategic purpose of the interface and the decision on a vector of the personality and tone of voice of the bot.
Paul Krizsan, Director Conversational AI

So while remaining up to date with the current developments of this exciting new technology is vital, we do not share the current ubiquitous sentiment that users are ready for unfettered access to potentially image-harming experiences without having some of the kinks of current LLMs ironed out over the course of 2023.

Are you interested in talking about conversational interfaces, LLMs and how to design for conversations? Talk to us!

Data-Driven Design–Designing with Data in a User-Centric Way

Posted on 5 August 202222 January 2026 by think moto

We talked to Marie Bossecker, Senior Experience Strategist at think moto, about Data-driven Design. She has many years of experience in combining data, strategy and design in such a way that they form the basis for development processes for digital products and services. We asked Marie what data-driven design actually is, how data-based design and creativity are connected, and how innovation gains quality through user data.

Data-driven design is radically user-centric and derives from design thinking. The first step, even before the strategy and design process begins, is an extensive collection of real user data that reflects the current behavior of the user group. Together with further data collection during the process, they form the basis for the development of new approaches in strategy and design.

“Data-driven design means making design decisions based on prior research and data analysis.”

The term data encompasses both the results from qualitative research, such as interviews, and quantitative research, such as surveys or tracking data.

What is the Data-driven Design process?

As with many design approaches, there is no clear process template. The steps presented here are a framework that can be used as is or modified slightly. As a structural basis, the 5 steps of Design Thinking serve: Empathize, Define, Ideate, Design and Test.

1. Data collection & analysis

Data can be collected using various methods, e.g., qualitative user interviews or quantitative data collection. Tools that anonymously query or record user behavior, such as in-page surveys, heat and click maps, or eye tracking, can be used for this purpose.

Data analysis is the task of the strategists. They interpret the data and filter out the problematic interfaces. Many modern tools for data analysis can help to identify conspicuous features and hierarchies that promote or negatively influence the performance of a website.

2. Definition

The task now is to react to the findings and assumptions made. This phase is accompanied by extensive research and, if necessary, user tests to re-examine the assumptions. There are many inclinations in the market that can affect user behavior. For example, the pandemic. User behavior has changed extremely as a result. These external influences and trends are highlighted and analyzed in the definition phase.

“Does a better conversion rate mean we’ve had success, or are there perhaps other movements in the market or in the target group that are influencing this result?”

3. Strategy

In this stage of research, a strategy/concept is developed based on the previous steps, which addresses the identified problems and includes possible solutions. As a rule, several approaches are developed here, which must prove themselves in the course of the further process or are just discarded.

4. Design & Implement

Based on the strategy, conceptual and design measures result, which are implemented by the designers. These are then implemented in the existing website. But the job is not done after that.

5. Test

After implementation, a test phase is carried out again to check how successful a measure was. The data obtained can then be used in turn to draw lessons and develop a revised strategy. This cycle is also known as “customer journey optimization”.

What role do strategists play in the data-driven design process, and how do they differentiate themselves from data analysts?

In quantitative methods, data analysts are primarily responsible for enabling data collection, i.e., creating an interface between the platform and the analysis tool, storing the data, and making it available to strategists in accessible dashboards. Interfaces, such as Google Analytics, hotjar or VWO, make the collection and transmission of data possible in the first place. In order to better evaluate the generated data, it is translated into dashboards and presented in an understandable way using data visualization. The strategists gain access to the data and can now evaluate it. Their task is to analyze and interpret the collected data, define measures, accompanied by extensive research, and then develop a strategy.

In qualitative methods of data collection, for example interviews or focus groups, strategists can be involved from the beginning. They develop the study, define aims and set the framework. After data collection, they then also evaluate the data.

What is the added value from combining strategy and data analysis in the design process?

With the flood of digital offerings, those who know their users best and create the best experience for them will prevail. The short attention span of users has made it all the more important to present relevant content in the most accessible way possible. The better the experience is tailored to the user and their needs, the longer their stay and the higher the likelihood of a “conversion,” such as a purchase or download.

It is almost impossible for designers today to include all the needs of potential user groups in design decisions. Some use the website very frequently, others only drop by occasionally. There are digitally affine personalities and those who need more assistance. That’s why it’s important for designers to draw on previous, data-based research. These show the current, real-world behavior of active user groups.

“You can’t know as a designer what your users really do or need without prior, data-based research. That’s where the clear difference lies between having some opinion and having some knowledge.”

Where does our Branded Interactions design process link to the Data-driven Design approach?

Data analysis can be well integrated in all phases of the branded interactions design process. It depends on the project and the industry of the customer how intensively the analysis of user data can be applied. Data collection is particularly helpful on websites with high traffic, where many users come together, such as in a large e-commerce store. Chatbots and their interfaces also provide a good basis for increasing performance through data in the long term. Qualitative data collection, on the other hand, can also support pure branding projects and MVBs and help to better understand the user group from the beginning through interviews and other research methods.

“Especially in the first two phases, Discovery and Define, data-driven strategy can be linked to the Branded Interactions design process. In Phase 5, Distribute, likewise, as the goal then is to evolve what has been implemented.”

Doesn’t creative freedom get lost if you always refer to data?

Real user data should not be a restriction on design freedom, but should serve as a support in the development of new design approaches. The data shows designers which approaches are already working well and which are not working at all. This allows them to focus on the essential pain points and create solutions where they are really needed. There are no limits to creativity itself.

Continuous analysis of user behavior helps us to optimize what we already have and adapt it to users in the best possible way. In order to develop new, innovative approaches, you have to keep questioning your previous knowledge to see what might work even better. Innovative design approaches can also be improved again and again through user testing and research.

What challenges do trends and technologies from the fields of tracking and data analysis bring for the combination of data and design?

In addition to external factors, such as pandemics, climate change or sustainability, current trends play a decisive role in how we behave online. For example, video content currently works much better than static content, as platforms like TikTok or Instagram guide. The need to be treated as an individual also has an impact on what we demand from our online experiences.

“When it’s my birthday, I expect a fat voucher from the brand I’ve already left hundreds of euros with.”

The line between personalizing content and manipulating buyers can be very thin. Every click and every text written reveals more about what we like and even how we feel right now. In parallel to the real personality, we also have a virtual one, which analytics tools build from our behavior, our data, and then feed us the content that best suits us.

“I believe that in the future, the line between manipulation and personalization will become narrower. The question is, after all, where do we draw the line? What is exploitation, what is convenience? As designers, we have a supporting responsibility to position ourselves.”

Want to learn more about the design process at think moto? You can read all about it in the book Branded Interactions by our founders. Also check out our project portfolio on thinkmoto.com to learn more about our work.