Want a free comprehensive digital product evaluation? Find out more

Using Conversational UX principles in your generative AI-powered products and services


In our blogpost about the changing design skills required to make generative AI-powered products more human and trustworthy, we suggested following the principles of Conversational UX and referenced a UX Magazine article by Linus Ekenstam called "7 Principles of Conversational Design"

Here are Ekenstam's seven principles seen through the lens of Generative AI to focus on UX techniques that build empathy and trust into your designs.

1. Engage: Be responsive and poignant. Reply if not interrupting.

  • Empathy & Trust: Quick, helpful responses demonstrate that you value the user's time. Generative AI can craft warm, personalized replies instead of generic templates. This conveys respect and makes the user feel heard.
  • Example: A customer struggling with setup gets an immediate response: "I understand this can be frustrating. Let's tackle this together step-by-step."

2. Recall: Be aware of context. Listen, don't ask things you already know.

  • Empathy & Trust: Remembering past interactions shows you care about the user's journey. Generative AI can prevent frustrations of repeating information by drawing from conversation history or stored personal preferences to offer tailored support. A great onboarding flow and progressive profiling can help build useful zero-party data that makes AI responses useful and in the right context.
  • Example: "I see you were looking at hiking backpacks last week. Did you have any specific questions about those models, or would you like to explore some other options?" or "You've told us before that sustainability matters to you; try these more eco-friendly choices."

3. Anticipate: Be predictive when possible.

  • Empathy & Trust: Anticipating needs demonstrates understanding and proactiveness. Generative AI can suggest relevant information, resources, or the next best action, making the user feel supported and guided.
  • Example: "Since you're interested in healthy meal planning, here are some recipes that might align with your dietary preferences." or "Do you want me to put the ingredients for that recipe into your weekly shopping list?"

4. Adapt: Conversation is organic, it flows, it has a rhythm. It's not mechanistic or rule-based.

  • Empathy & Trust: Being flexible and accommodating to conversational shifts signals you're listening, not just following a script. Generative AI can adapt to changes in tone, understand nuances, and even respond appropriately to humor or sarcasm, building rapport.
  • Example: "That's a great point about accessibility! Let's explore some alternative design options that would be more user-friendly for everyone."

5. Reflect: Repeat important pieces of information to confirm user needs.

  • Empathy & Trust: This active listening technique builds trust by ensuring alignment. Generative AI can summarize complex instructions or decisions, allowing the user to verify their understanding and feel confident in moving forward.
  • Example: "So, you'd like to schedule a follow-up appointment for next Tuesday at 2 PM. Is that correct?"

6. Pull: Use cues, directions, or instructions to steer the flow of the conversation.

  • Empathy & Trust: Providing gentle guidance shows you're committed to helping the user reach their goals. Generative AI can offer well-timed suggestions or questions without taking away the user's sense of control, fostering a collaborative experience.
  • Example: "It seems like you're feeling overwhelmed with options. Would you like me to narrow the search results further, or would you prefer to explore on your own for a while?"

7. Flow: If your service operates asynchronously, send users a simple message if no one is online. If your service is synchronous (powered by a bot/machine) don't bombard them when they send their first message.

  • Empathy & Trust: This principle is about setting expectations and respect. Generative AI can manage asynchronous interactions gracefully, offering updates or options to leave contact information, showing the user they haven't been forgotten. In synchronous interactions, starting with a simple greeting and asking how to help builds trust rather than overwhelming the user.
  • Example: "Hello! I'm your virtual assistant. How can I help you today?"

Get in Touch Potato helps companies to build digital products and services that are purposeful and beautifully designed. If you want to discuss how to use generative AI to make your experiences for customers more conversational and compassionate, give us a call.