Defining the Problem
Situation
Convo helps researchers scale user research by conducting multiple AI-facilitated qualitative interviews concurrently, saving time and manpower costs.
Key Interviewee Pain Points (from research):
Interviewees have surfaced concerns about:
  • Concerns around privacy
  • Confusing navigation
  • Distractions from transcript errors
  • A lack of human connection, making the AI feel cold or robotic
“How might we make AI interviews feel less robotic and more human?”
  • Address pain points in the user journey map
  • Features that engage Gen Zs
Affinity Mapping
Interviewed 7 participants (ages 25–45, familiar with AI) to evaluate onboarding, interview flow, and exit experience. Focused on pain points around privacy, navigation, and emotional comfort, which informed design improvements to build trust and humanise the AI interaction.
The User Journey Flow (Before Redesign)
There are three main screens in the interview process.
Original Screens (Onboarding, Interview, and Completion)
User Journey Map 👣
Synthesised interview findings from the affinity map to uncover pain points, emotions, and design opportunities throughout the user journey — from onboarding to completion.
👤 User Persona
Tech-savvy Thomas
Novelty
Control
Social Connection
"I value a clear purpose, judgment-free conversation, and control over how I share my thoughts."
Phase 1: Onboarding
Key Actions:
Fill in personal details
Choose an interview type
What users said:
  • “Who’s the one looking at the transcripts?”
  • Choose an interview type (Chat vs Voice)
  • “Am I supposed to choose a certain type here?”
  • “What is the difference between Chat and Voice interview?”
How users felt:
😰
😕
Pain Points:
Users are concerned about data privacy and want to feel in control regarding what data they share.
No clear CTA
Opportunity:
  • Inform users that they can end the interview at any time
  • Explain where the data will be recorded and how it will be used
Provide a walkthrough of the Voice and Chat buttons for clarity
Phase 2: Interview
Key Actions:
Listen to & answer interview questions
End interview if uncomfortable
What users said:
  • “Oh it started”
  • “If transcription is wrong I don't know how to correct it or should I correct it.”
  • “There's no feedback to me in the sense if the AI takes in what I say...:”
  • “They keep cutting me off”
  • “This feels robotic and one-sided compared to talking to a human.”
  • “Oh I didn’t think I could end the interview”
  • “What the AI records, ask them to press the end interview if they are not comfy”
How users felt:
😧
😢
Pain Points:
  • Users are startled by the abrupt interview start
  • Live transcription is distracting and makes users self-conscious
  • Users want clearer system feedback
  • Overall emotional disconnect
“End interview” button does not improve users’ comfort, induces some anxiety instead
Opportunity:
  • Show animated voice lines that move when the user talks to indicate the AI is listening
  • Remove transcription feature
  • Use a static face tied to the voice to humanise the interaction
  • Let users know they can ask Convo to pause
Provide a walkthrough of the Voice and Chat buttons for clarity
Phase 3: Interview Completion
Key Actions:
Recognise that interview has ended, have a good impression of Convo
What users said:
  • “Oh that’s all?”
  • “At the end of the interview, user is unsure do I close the screen, do I go back to the website or submit another response?”
How users felt:
😕🤔
Pain Points:
  • The ending page falls flat, and does not impress
  • Users are unsure on what to do after completing the interview
Opportunity:
  • Add some form of appreciation
  • Recommend adding a CTA to guide users on what to do next e.g. They can close the window or redeem the voucher incentive
Wireframe Sketches
Sketches explored the interview flow across mobile and desktop, with early thoughts on user guidance, visual hierarchy, and tone.
Low-Fidelity Wireframes
Mobile
Desktop
Prototype (Final Design)
Visual UI elements like learning beads were used to reinforce the platform’s role as a learning-driven interview tool, while emojis created a friendly, relatable tone with users.
Mobile
Desktop
The Revised User Journey Flow
Results
PortfolioFish Mart SAKURAYAAirAsiaAmretConvo
CV/Resume
8541 5621
© 2025 Emmeline. All rights reserved.