50% Voucher
Professional man using a chat based UI on his phone.

Conversation design is the emerging field of crafting natural and intuitive dialogues between humans and digital systems, shaping how we interact with technology through speech and text.

The idea of having a conversation with your computer isn’t new. For decades it’s been played out in sci-fi movies. And in a sense, the early computer interfaces where a kind of dialog – a text prompt where you could type commands and where the computer could also ask you questions

Key Takeaways



Persona Nuances for Conversational UX Personas for conversational UX need to consider conversational contexts, multimodal preferences, and cultural nuances to reflect natural language interactions.
Real World Conversational UX Design Tips Tips include keeping experiences focused, writing for spoken language, managing user expectations, and designing for context and multimodality.
Multimodal Value and Constraints Incorporating multimodal abilities can enhance complex tasks, accessibility, and context-aware experiences, but should be balanced to avoid unnecessary complexity.
Anthropomorphic Design Considerations Effective anthropomorphic design involves balancing human-like appearance and interaction, considering ethical implications, and using multimodal cues.
Conversational UX in Action Examples include Bank of America’s Erica, Duolingo, Domino’s Pizza, Skyscanner, and Sephora, demonstrating the practical application and benefits of conversational interfaces.
Avoiding Conversational UX Pitfalls Key strategies include defining clear use cases, planning for human-agent handover, using rich conversational elements, and implementing robust testing and guardrails.
Future of Conversational UX Future possibilities include multimodal experiences, personalized agents, emotionally intelligent interactions, and ubiquitous conversational interfaces.

The Rise of Conversational Interfaces

The vision of conversational interfaces dates back to the 1960s with early research on dialogue systems and text-based chatbots for question answering. However, these initial efforts were limited by the computational power and natural language processing capabilities of the time.

Over the following decades, advancements in artificial intelligence and language models gradually improved the sophistication of conversational interfaces. Major tech companies like Google, Amazon, Microsoft, and Apple pioneered voice-based conversational assistants like Google Assistant, Alexa, Cortana, and Siri, leveraging improved speech recognition and language understanding models.

The recent breakthrough of large language models like GPT-3 and its successor ChatGPT has ushered in a new era of conversational AI, enabling more natural, context-aware, and open-ended dialogues. ChatGPT’s ability to engage in human-like conversations, provide detailed responses, and even generate code or creative content has captivated users and demonstrated the potential of conversational interfaces to revolutionize how we interact with technology.

While challenges remain, such as ensuring safety, accuracy, and ethical use of these powerful language models, the rapid rise of ChatGPT and similar systems has reignited interest and investment in conversational AI, paving the way for more advanced and ubiquitous conversational interfaces in the near future.

Contrasting Traditional and Conversational UX

Here are the key differences between UX design for traditional human-computer interfaces and Conversational UX:

Interaction Model

Traditional UIs: Rely on direct manipulation through graphical elements like menus, buttons, and forms. Users initiate actions and the system responds.

Conversational UX: Based on natural language dialogues, mimicking human-to-human conversations. Users express intents through speech or text, and the system interprets and responds accordingly.

Information Architecture

Traditional UIs: Information is structured hierarchically with menus, pages, and navigation paths. Users follow predefined flows.

Conversational UX: Information is organized around conversational topics and intents. Users can jump between contexts more fluidly.

Input Modalities

Traditional UIs: Primarily rely on keyboard, mouse, and touch inputs.

Conversational UX: Supports multimodal inputs like voice, text, gestures, and even gaze or facial expressions.

Output Modalities

Traditional UIs: Primarily visual outputs through graphical user interfaces.

Conversational UX: Can leverage multiple output modalities like voice, text, visuals, and even haptic feedback.

Context and State Management

Traditional UIs: Context and state are managed through explicit user actions and system responses.

Conversational UX: Context and state must be inferred from natural language inputs, requiring advanced language understanding and dialogue management.

Error Handling

Traditional UIs: Errors are typically handled through explicit error messages and recovery paths.

Conversational UX: Error handling requires more natural language generation and conversational repair strategies.

Personalization and Adaptation

Traditional UIs: Personalization is often limited to customizing visual elements or preferences.

Conversational UX: Conversational agents can adapt their language, personality, and knowledge based on individual user preferences and interaction history.

While traditional UIs and conversational UX share some fundamental design principles, the shift towards natural language interactions introduces unique challenges and opportunities in areas like context management, multimodal input/output, and creating more human-like conversational experiences.

Persona Nuances for Conversational UX

When creating conversational UX compared to traditional UX design, defining user personas and the target audience follows similar principles but with some key differences:

Conversational Contexts and Intents

In conversational UX, personas need to capture the different conversational contexts, intents, and language patterns users might employ when interacting through natural dialogue. This goes beyond traditional task flows and encompasses the various ways users might phrase requests or queries.

Multimodal Preferences: Conversational UX personas should account for users’ preferred input and output modalities (e.g., voice, text, visuals) and how they might combine these modalities during interactions. Traditional UX personas primarily focus on graphical interface preferences.

Personality and Rapport: Since conversational UX aims to mimic human-like dialogues, personas may need to reflect desired personality traits, communication styles, and the level of rapport or familiarity users expect from the conversational agent. This is less critical for traditional GUI-based interfaces.

Language and Cultural Nuances: Conversational UX personas must consider language nuances, idioms, dialects, and cultural contexts that can influence how users communicate and interpret responses. This is especially important for multilingual or global user bases.

Error Handling and Recovery: Personas should capture users’ expectations and preferences for error handling and conversational repair strategies when the system fails to understand or respond appropriately. This is a unique aspect of conversational UX not present in traditional UIs.

Contextual Awareness: Conversational UX personas need to reflect users’ expectations around the system’s ability to maintain context and state across multiple conversational turns, as well as their tolerance for context switching or topic changes.

While traditional UX personas focus on user goals, tasks, and interface preferences, conversational UX personas must additionally account for the nuances of natural language interactions, multimodal preferences, personality traits, and contextual awareness to create truly human-centric conversational experiences.

Real World Conversational UX Design Tips

Here are some tips from real-world conversational UX designers:

Keep it Simple and Focused

Conversational experiences should be designed around specific use cases and user goals. Avoid trying to create an all-encompassing conversational agent that attempts to handle every possible scenario. Focus on solving core user needs effectively.

Write for Spoken Language

Conversational scripts should mimic natural spoken language patterns, using contractions, colloquialisms, and a conversational tone. Avoid overly formal or robotic language. Test dialogues by reading them aloud to ensure they sound natural.

Manage Expectations

Set clear expectations about the capabilities and limitations of the conversational interface. Don’t overpromise or create unrealistic expectations that could lead to user frustration. Provide guidance on how to interact effectively.

Design for Context

Conversational UX should account for the user’s context, such as their location, device, previous interactions, and the current situation. Tailor responses and information accordingly for a more relevant and personalized experience.

Embrace Multimodality

Incorporate multiple input and output modalities like voice, text, visuals, and gestures to create more engaging and accessible conversational experiences. Leverage the strengths of each modality for different interaction types.

Prioritize Error Handling

Robust error handling and conversational repair strategies are crucial. Provide clear feedback when the system doesn’t understand, offer alternative paths, and allow users to rephrase or restart the conversation easily.

Iterate and Test

Continuously test and refine conversational flows, language models, and responses based on real user interactions and feedback. Conversational UX is an iterative process that requires ongoing optimization.

Collaborate Across Disciplines

Conversational UX design requires collaboration between UX designers, writers, linguists, voice interaction experts, and developers. Leverage diverse perspectives to create cohesive and effective conversational experiences.

By following these tips from experienced conversational UX designers, teams can create intuitive, engaging, and user-friendly conversational interfaces that meet user expectations and deliver value.

Multimodal Value and Constraints

Incorporating multimodal abilities in conversational UX can add significant value in certain scenarios, while in others, it may be more effective to focus on one or two primary modalities. Here are some considerations:

When Multimodal Abilities Add Value:

  1. Information-Dense Tasks: For tasks involving complex information, visuals, or data visualization, combining voice/text with graphical elements can enhance understanding and make the experience more intuitive. Examples include research, data analysis, or product customization.

  2. Multitasking Scenarios: When users need to engage with a conversational interface while performing other tasks, multimodal interactions allow them to switch between voice, text, and visuals seamlessly. This is valuable for scenarios like driving, cooking, or working on a computer.

  3. Accessibility Needs: Offering multiple input and output modalities caters to users with diverse abilities and preferences, improving accessibility and inclusivity. Multimodal interfaces can accommodate visual, auditory, motor, or cognitive impairments.

  4. Context-Aware Experiences: Combining modalities like voice, visuals, and location data can create more contextually aware and personalized conversational experiences tailored to the user’s situation.

When Focusing on One or Two Modalities is Better:

  1. Simple, Straightforward Tasks: For basic queries, commands, or transactional tasks that don’t require extensive information exchange, a single modality like voice or text may suffice. Multimodal interactions could overcomplicate the experience unnecessarily.

  2. Mobile or Hands-Free Scenarios: In situations where users are on the go or have limited ability to interact visually, focusing on voice or text input/output can provide a more streamlined and convenient experience.

  3. Limited Device Capabilities: If the target devices have constraints in terms of screen size, processing power, or input/output capabilities, it may be more practical to prioritize one or two modalities that work well within those limitations.

  4. Consistency and Familiarity: In some cases, users may prefer a consistent experience across devices or platforms, making it more suitable to stick with a familiar modality like voice or text rather than introducing multimodal interactions.

Ultimately, the decision to incorporate multimodal abilities or focus on one or two modalities should be driven by the specific use case, user needs, context, and device capabilities. Striking the right balance between modalities can enhance the conversational UX, but unnecessary complexity should be avoided.

Anthropomorphic Design Considerations

Anthropomorphic design involves imbuing non-human entities with human-like qualities, forms, or behaviors. When applied to robots and conversational interfaces, key considerations include:

Measuring humanness in appearance and interaction separately, then evaluating their alignment. Appearance humanness captures visual anthropomorphism, while interaction humanness reflects human-like behaviors, responses, and capabilities. Harmonizing these two aspects prevents dissonance between a robot’s human-like appearance and limited interactivity, or vice versa.

Determining the appropriate degree of anthropomorphism based on the robot’s intended role and context. Excessive anthropomorphism can raise unrealistic expectations, while insufficient human-likeness may hinder acceptance and emotional connection. A balanced approach tailored to the specific use case is recommended.

Considering multimodal anthropomorphic cues beyond just visual form, such as gestures, speech patterns, emotional expressions, and personality traits. Coordinating these modalities enhances the perception of human-likeness and emotional engagement.

Addressing ethical implications as human-robot relationships grow closer. Discussions around privacy, moral accountability, and potential deception highlight the need for responsible anthropomorphic design practices.

Overall, anthropomorphic design requires carefully calibrating human-like qualities across multiple dimensions while managing user expectations and ethical considerations for optimal human-robot interaction experiences.

Conversational UX in Action

Here are some real-world examples of Conversational UX:

  1. Bank of America’s Erica: Erica is a virtual assistant chatbot that helps Bank of America customers with basic banking tasks like checking account balances, transferring money, and getting account information. With over 19.5 million users, Erica has improved customer service and enabled conversational banking experiences.

  2. Duolingo: The popular language learning app Duolingo uses conversational exercises, voice recordings, and chatbots to create an interactive experience of learning languages through simulated conversations. Its conversational approach has helped Duolingo gain over 500 million registered users.

  3. Domino’s Pizza: Domino’s Pizza allows customers to place orders, customize them, and track deliveries through a conversational chatbot on Facebook Messenger. This conversational interface streamlines the ordering process for a better user experience.

  4. Skyscanner: The travel search engine Skyscanner has integrated conversational chatbots on platforms like Telegram and Skype, allowing users to search for flights and hotels through natural language conversations. Skyscanner has also enabled voice interactions with Amazon’s Alexa for booking flights.

  5. Sephora: The cosmetics brand Sephora uses a conversational chatbot on messaging apps like Kik to provide personalized product recommendations, makeup tutorials, and assistance with purchases based on user preferences and conversations.

These examples showcase how conversational UX powered by chatbots, voice assistants, and AI-driven dialogues is being adopted across various industries to enhance customer experiences and enable more natural human-computer interactions.

When Conversational UX Goes Wrong

A cautionary tale of conversational UX gone wrong emerged when a car dealership’s ChatGPT-powered chatbot was exploited by pranksters, leading to viral incidents of the bot agreeing to sell vehicles for $1 and engaging in off-topic conversations.

The chatbot, implemented by Fullpath for Chevrolet of Watsonville, was meant to assist shoppers but lacked robust guardrails, allowing users to coax it into making legally dubious claims and straying from its intended automotive focus.

While Fullpath defended the bot’s resistance to most pranks, the viral examples highlighted risks of conversational AI lacking proper constraints and context awareness, underscoring the need for rigorous testing and fail-safes in deploying conversational interfaces.

Avoiding Conversational UX Pitfalls

To avoid pitfalls when designing conversational UX while ensuring effectiveness, consider the following:

  1. Clearly define the use case and goals for the conversational experience: Outline specific problems it will solve and tasks it will enable. Without a clear purpose, the development becomes unfocused.

  2. Plan for seamless human-agent handover when interactions cannot be fully automated: Ensure a direct path to human assistance for complex cases to maintain user satisfaction.

  3. Leverage rich conversational elements like graphics, carousels, and voice input: Text-only chatbots often fail to create engaging experiences. Utilize multimodal interactions for simplicity and context.

  4. Build for omnichannel deployment and scalability from the start: Aggregate messaging channels to manage context, handoffs, and integrations as conversational volume grows.

  5. Avoid overcomplicating the experience with unnecessary AI capabilities: Focus on core functionalities that truly enhance the user experience rather than gimmicky features.

  6. Continuously test and refine the conversational flow, language models, and failure handling: Rigorous testing helps identify potential issues before public deployment.

  7. Implement robust guardrails, context awareness, and fail-safes: This prevents the conversational AI from straying off-topic or making inappropriate responses that undermine trust.

By following these principles, conversational UX designers can create effective, engaging experiences that solve real user needs while mitigating common pitfalls that could hinder adoption and satisfaction.

Conversational UX Limitations

While conversational UX offers an intuitive and natural way to interact with digital systems, it is not universally suitable for all use cases. Scenarios that require precise control, complex data manipulation, or visualizing large amounts of information may be better served by traditional graphical user interfaces (GUIs).

For example, tasks like coding, financial modeling, or data analysis often benefit from the structured layout and direct manipulation afforded by GUIs. Additionally, conversational interfaces can struggle with ambiguity, context-switching, and handling complex multi-step workflows, making them less ideal for certain productivity applications. As such, conversational UX should be selectively applied where it genuinely enhances the user experience, rather than forcing it as a one-size-fits-all solution for human-computer interaction.

Accessible Conversational UX Design

To ensure conversational UX is accessible, consider the following:

  1. Provide multimodal input and output options: Allow users to interact through speech, text, or a combination of both. Offer alternative output modes like text transcripts, visual aids, or audio descriptions to accommodate different needs.

  2. Implement clear and concise language: se simple, unambiguous language that is easy to understand, avoiding complex jargon or idioms. This aids users with cognitive impairments or limited language proficiency.

  3. Support assistive technologies: Ensure compatibility with screen readers, speech recognition software, and other assistive tools. Follow accessibility standards like WCAG for proper markup and labeling.

  4. Allow customization and personalization: Enable users to adjust settings like text size, contrast, speech rate, and volume to suit their preferences and needs. Personalized profiles can store these settings.

  5. Implement error handling and recovery: Provide clear error messages and guidance when the system fails to understand user input. Allow users to easily restart or rephrase their queries.

  6. Conduct accessibility testing: Involve users with diverse abilities throughout the design and testing process. Gather feedback on potential barriers and iterate to improve accessibility.

  7. Comply with accessibility regulations: Adhere to relevant accessibility laws and guidelines, such as the Americans with Disabilities Act (ADA) or the Web Content Accessibility Guidelines (WCAG).

By incorporating these practices, conversational UX designers can create inclusive experiences that cater to users with varying abilities, ensuring equitable access and usability for all.

Future of Conversational UX

Here are some exciting possibilities for the future of Conversational UX:

  1. Multimodal Conversational Experiences: Future conversational UX will seamlessly blend multiple modalities like voice, text, visuals, and gestures for more natural and immersive interactions. Users could converse with AI assistants while simultaneously sharing visual information or using gestures to provide context.

  2. Personalized Conversational Agents: Conversational agents will become highly personalized, adapting their language, personality, and knowledge to individual users based on their preferences, context, and interaction history. This could create a sense of familiarity and rapport, enhancing user engagement and trust.

  3. Emotional Intelligence and Empathy: Advancements in affective computing and natural language understanding will enable conversational UX to better recognize and respond to user emotions, providing empathetic and emotionally intelligent responses. This could improve user satisfaction, especially in domains like customer service or mental health support.

  4. Conversational Commerce and Transactions: Conversational UX will facilitate seamless commerce experiences, allowing users to discover, research, and purchase products or services through natural conversations with virtual assistants or chatbots. This could streamline the buying process and enhance customer experiences.

  5. Ambient Computing and Ubiquitous Conversational Interfaces: With the rise of ambient computing and the Internet of Things, conversational UX will become ubiquitous, enabling users to interact with devices and systems through voice or text in any environment, from smart homes to connected vehicles.

  6. Conversational Accessibility and Inclusivity: Conversational UX has the potential to improve accessibility and inclusivity for users with disabilities or language barriers, enabling them to interact with technology more naturally and intuitively through speech or text.

As conversational AI and natural language processing technologies continue to advance, the possibilities for conversational UX are vast, promising to revolutionize how we interact with digital systems and integrate them into our daily lives.

Conversational Design with resonio

Tools like resonio can be helpful in designing conversational interfaces in several ways:

  1. User Research and Feedback: resonio’s survey creation tools and access to a large pool of international participants enable gathering valuable user insights, preferences, and feedback during the design process. This data can inform the conversational flow, language models, and personas to create more natural and user-centric conversational experiences.

  2. Prototype Testing: resonio’s real-time tracking and reporting features allow designers to test early prototypes of conversational interfaces with target users, identifying areas for improvement and iterating based on user interactions and feedback.

  3. Usability Evaluation: Conducting usability studies through resonio can help assess the intuitiveness, effectiveness, and accessibility of conversational interfaces, identifying potential pain points or areas of confusion for users.

  4. Continuous Improvement: resonio’s ability to rapidly deploy and gather feedback on surveys enables ongoing optimization of conversational experiences based on evolving user needs and preferences.

By leveraging resonio’s market research capabilities, designers can gain valuable insights into user behaviors, expectations, and pain points, enabling them to create more human-centric, intuitive, and effective conversational interfaces that resonate with their target audiences.


What is the primary difference between Conversational User Interfaces (CUI) and Voice User Interfaces (VUI)?

Conversational User Interfaces (CUI) include both text- and voice-based interfaces where users communicate in natural language. Voice User Interfaces (VUI), on the other hand, are part of the CUI concept and imply voice-only interaction with devices.

How do you ensure a consistent persona in a conversational interface?

To ensure a consistent persona, define a target user's personality traits and maintain consistency in the tone, words, and phrases used by the conversational UI. This helps in creating a natural and authentic conversational experience.

What are the key steps in designing a conversational flow?

Designing a conversational flow involves writing dialogues for different scenarios, such as onboarding, sales, and follow-up interactions. This helps in understanding how the conversation might unfold and ensures a natural flow.

How do you adapt conversational UX to different business objectives?

The choice of a chatbot's goals and tasks depends on the company's business objectives. For instance, a technical support bot might aim to reduce costs, while a lead generation bot might focus on increasing profits. Understanding the target audience and their needs helps in crafting helpful responses that align with the business objectives.

What is the importance of user research in conversational UX?

User research is crucial in conversational UX as it helps in understanding the user's pain points and preferences. Conducting user interviews and gathering qualitative data ensures that the conversational interface is designed to meet the user's needs effectively.