Step 1: Provide Task Details for the Participant
This step focuses on giving participants clear instructions about the tasks they should complete. These details are crucial for helping the AI Agent guide them accurately throughout the interview.
Create a Clear Task Title and Description
Clearly define what the participant should do on each screen.
The AI Agent relies on this information to deliver meaningful guidance.
Keep instructions simple, direct, and concise to minimize confusion.
Provide a Focused, High-Level Task description
Describe the overall objective the user should accomplish. This could be testing a feature, navigating a screen, or completing a specific action.
The description is visible to the participant during the session, helping them stay oriented and understand what they need to accomplish.
Step 2: Include Your Figma Prototype Link and Screen Topics
Properly configuring your prototype ensures smooth AI performance and accurate participant guidance. Follow these best practices when setting up your Figma file.
Limit Figma Interactions to a Maximum of 5 Per Screen
Keep the number of interactions (clicks, drags, or gestures) under five per screen to help the AI process inputs effectively and maintain conversational clarity.Use Clear and Unique Names for Frames
Each frame in Figma should have a distinct, descriptive name so the AI agent can differentiate between screens easily.Keep Screens Short and Simple
Avoid long, scrolling screens or unnecessary visual clutter. Simple screens make it easier for the AI to track user actions and provide precise feedback.Limit the Number of Screens to 10
The AI can process up to 20 total images, including user interactions. Limiting your prototype to 10 screens allows room for up to 10 user interaction states, optimizing AI performance.Add Specific Topics or Questions for Each Screen
If you want the AI Agent to ask specific questions on particular screens, define those prompts clearly.
Example:“Ask the participant what factors influence their decision to click this button.”
“Ask how easy it was to find the ‘checkout’ option.”
⚠️ Note: Once your Figma file is uploaded to Genway, any subsequent changes made in Figma will not be automatically reflected in Genway.
These focused questions ensure the AI gathers targeted feedback at the right time.
Step 3: Define Overall Interview Guidelines
This step defines the broader interview guidelines - a cheat sheet for the AI to follow during the test. These help the AI stay aligned with your objectives and handle unexpected user behavior.
Provide Overall Guidelines, Not Detailed Instructions
Use this section to describe the purpose and flow of the interview. Avoid including detailed screen-level questions here, those belong in Step 2.Focus on the Key Objectives of the Interview/ Highlight what success looks like
For example: discovering usability pain points or evaluating navigation efficiency.Use Guidelines as a “Cheat Sheet” for the AI
Add context and helpful cues for the AI Agent to use during the conversation. Example:“If the participant seems stuck, encourage them to explain what they’re looking for.”
“If they complete the task quickly, ask how intuitive it felt.”
Monitor and Adjust Based on AI Performance
After running tests, review how well the AI tracked interactions and followed your guidance. Refine your instructions or guidelines to improve future sessions.
Step 4: Review and Iterate
After completing a few sessions, revisit your Figma setup, questions, and guidelines. Identify where participants struggled or where the AI needed more context, and refine accordingly.
Thank you for following this guide, and happy testing!