Case Study - Deep talks

The problem

In a world where we’re more connected than ever, it often feels like we’re becoming more distant. Social media, trends, and the fast pace of modern life have shifted us toward quick, shallow interactions. The art of deep conversation, once essential, now feels rare.

And that’s a problem.

Meaningful conversations are powerful—they build trust, strengthen relationships, and help us understand each other better. But starting and maintaining them can be difficult. Why is that?

With this in mind, I set out to design a digital product that makes meaningful conversations easier and more natural. A tool that doesn’t just ask questions, but genuinely guides deep, reflective dialogue—fueling the human need for connection and fostering stronger bonds with others and ourselves.

Understanding the problem: Pain Points

I started by interviewing friends and acquaintances about the topic, with the following research goals in mind:

• Understanding the context and motivations behind deep conversations
• Understaing the frustrations they may bring

From these interviews, I identified four key pain points that people face when starting or maintaining deep conversations:

1. Breaking the ice – Initiating a deep conversation can feel intimidating.
2. Managing discomfort – Some topics can quickly become heavy or awkward.
3. Vulnerability hesitation – Many are afraid to open up.
4. Shyness – Sometimes it’s hard to know what to say next.

Persona development: Meet Marina

Based on these pain points and a deeper understanding of the struggles in this situation, I focused on designing a user persona that would enhance empathy and allow me to analyze the context, desires, and pain points more clearly.

I made sure the user persona captured:

• A relevant quote that conveyed the persona’s attitude and personality.
• A list of goals the persona aims to achieve.
• A list of frustrations they experience or try to avoid.
• A brief scenario outlining goals, frustrations and other important context.

Based on this user persona, the problem to solve became much clearer. The resulting problem statement is as follows:

User journey

From all the insights gathered, the idea emerged: a platform designed to initiate deep and meaningful conversations in a comfortable way, offering curated and adaptive questions that encourage reflection, powered by an AI algorithm.

Mapping out the user journey helped pinpoint key touchpoints, uncover emotional experiences, and identify opportunities for improvement—laying the foundation for defining the features.

Competitive analysis

I explored many apps designed to spark deep conversations, but most were focused on couples or parties, with only a few fostering connections between friends, family, or oneself.

I also noticed that many apps used fixed, predefined question decks, causing repetition once the deck was completed—similar to a board game. These decks didn't adapt to users, and sometimes included uncomfortable topics.

Value proposition: Limitless, adaptive questions.

The deep talks app stood out by offering limitless questions powered by AI. It adapts based on user feedback, ensuring a comfortable, enjoyable conversation tailored to context and preferences, minimizing any potential discomfort.

The main features of the product are as follows:

• Preliminary questionnaire – Trains the AI on user preferences.
• Real-time adaptive questions – Users give feedback on questions, allowing the AI to refine them instantly.
• Conversation history – Users can save and revisit past conversations.
• Engagement with friends – A social feature encouraging shared experiences.

My proposal: User flow

Gaining insight into user interactions was essential in shaping the experience. The main journey developed as follows:

- Design -

Sketching

After collecting all the insights, I dove into the design process, keeping everything I had learned in mind—user personas, pain points, and the updated flows. I began visualizing the concepts, outlining initial ideas, and establishing the groundwork for the final design.

Lo-Fi Design

Using the wireframes, I created a Lo-Fi prototype that mapped out the main flow, from the homepage to starting a personalized conversation for the user.

Usability study

I conducted remote usability tests via Figma, tracking participant behavior while encouraging them to think out loud. Participants were asked to complete the following tasks:

Tasks:

• A group of 4 people who want to start a conversation: navigate until you start the questions
• Provide feedback on a question
• Save the conversation for later and exit
• Resume a saved conversation

Conclusions:

I concluded that participants didn’t fully understand that their feedback was actively changing the questions in real-time to adapt to them.

• To address this, I updated the copy for better clarity (before and after wireframes).
• This was a crucial feature, as users needed to understand that their feedback was directly influencing future questions, making the experience more personalized and comfortable—maximizing the product’s potential.

In addition to that:

• I changed the exit icon, as the door symbol wasn’t the most intuitive.
• For the multiple-choice questions in the quiz, I made sure it was clear which options had been selected.

Hi-Fi Design

For the high-fidelity design and interface, I aimed to convey a sense that conversations could be fun and interesting—not heavy or serious. I wanted it to feel dynamic, heartwarming, and casual, while also reflecting something modern and part of everyday life.

To achieve this, I used vibrant colors and a neo-brutalism style—modern and trendy—to bring a fresh, fashionable vibe to the conversation experience.

Homepage

Beyond starting a new conversation, this is the user's personal hub—a safe space designed just for them. From here, they can connect with friends, revisit past conversations, and track their progress toward having more meaningful interactions with the people who matter most to them.

Saved conversations

This feature lets users revisit past conversations for insights or enjoyment, seamlessly resuming where they left off. Saved chats also refine AI understanding, enhancing future interactions.

Questionnaire

Fill out a quick questionnaire to set up your conversation based on your preferences and basic details. Indicate whether you're chatting alone or with others, and specify the number of participants. You can also add contacts directly, allowing them to be registered within the conversation for a more personalized experience.

Topic selection

Customize your experience by selecting topics that interest you. This helps the AI refine its understanding, tailoring conversations to be more engaging, relevant, and aligned with your preferences. Over time, this leads to a more dynamic and personalized interaction, ensuring the AI provides more meaningful questions.

Question screen

Provide feedback on the questions as they appear so the algorithm can adjust them in real time, making each new question more relevant to your needs. You can also tweak settings like closeness level and general preferences directly from this screen. Plus, you have the flexibility to exit and save the conversation at any moment.

Some key UI considerations:

• To make certain navigational cues clearer, I added depth and shadow to interactive elements like buttons.
• I included a progress bar so users could track their progress in the quiz, reducing frustration and keeping them informed (one of the core principles of good design, according to Norman Nielsen's 10 UI principles).
• The question screen was designed to be as minimalist as possible, with the fewest elements—allowing users to focus on what matters most: the conversation they’re having.

Evaluation and reflecting

• Gaining a deeper confidence and fluency in applying the design thinking methodology has been one of the most rewarding parts of this process. Using tools like user personas, user journeys, and competitive audits has allowed me to approach design with a more structured and user-centered mindset. This has not only enhanced my skills but also reinforced the importance of truly understanding the user at every stage.
• I’ve also learned that iterations are essential to creating something truly impactful. It's easy to think that changes and adjustments are setbacks, but in reality, they are opportunities for growth and improvement. Paying close attention to how users interact with the product and how they react to the tasks I’ve set for them is key. Every tweak, every shift in direction, is a step toward a more refined, user-focused experience. This process has taught me that design is never about perfection from the start, but about evolving and adapting to better serve the people who use it.

Honed skills