This document provides a high-level overview of the participant experience during a Tellet interview.
Welcome & Consent — Duration: 1-2 min; Can Skip?: No; What Happens: Read intro, choose language, accept terms
Screening — Duration: Varies; Can Skip?: Shown only when the study uses screening; What Happens: Answer qualification questions if configured
Main Interview — Duration: Varies; Can Skip?: No; What Happens: Answer interview questions via text/speech/video
Completion — Duration: Immediate; Can Skip?: No; What Happens: View completion message, follow next steps
Total duration: Varies based on number of questions and AI probing configuration. General rule: ~45 seconds per question.
Participants land on a welcome screen that displays:
Company logo
Introduction message explaining the AI interview format
Language selection dropdown (if enabled)
Privacy Policy and Interview Terms links
Consent checkbox
"Let's go" button
If multiple languages enabled:
System automatically detects browser language
Participant can select different language from dropdown
All interface elements translate to selected language
Can pre-set via URL, using either the full language name or its 2-letter code: ?language=Spanish or ?language=es
How the starting language is chosen: if the link includes a ?language= parameter, that language is used; otherwise the participant's browser language is used (when language selection is enabled); otherwise the project's default language is used.
For languages that read right-to-left — Arabic, Hebrew, Persian, and Urdu — the entire interface is laid out right-to-left.
Requirements:
Participant must check the consent checkbox
Checkbox text varies by project configuration:
For recording-enabled interviews: "I consent to the recording of [recording type] in this interview and to sharing the recording with the research client for research purposes."
For text-only interviews: "I consent to participate in this interview and to sharing my responses with the research client for research purposes."
Cannot proceed without consent
After accepting:
System records consent with timestamp
Saves language preference
Creates unique session
Advances to next stage
Filter participants based on qualification criteria before the main interview.
Progress indicator shows "Question X of Y"
One question displayed at a time
Company logo visible
Clear question text
Single-select: Radio buttons, one answer only
Multiple-select: Checkboxes, multiple answers allowed
Numeric input: Text box for numbers
Screening questions are translated into the participant's language ahead of time, so there is no wait at this step.
If qualifies: Proceeds to next question, then main interview
If doesn't qualify: Sees custom disqualification message, interview ends. If a panel Screened Out link is configured, the participant sees a short countdown and is automatically redirected there, the same as on the completion screen, with the option to cancel or copy the link instead.
If a screening quota is full: Sees a separate quota full message instead (the participant qualified, but the quota for their answer was already reached). The same automatic countdown-and-redirect applies if a panel Quota Full link is configured.
Chat-style layout:
AI questions on left (neutral)
Participant responses on right (coloured)
Progress indicator at top
Input options at bottom
Participants can respond using any combination of:
Text — How It Works: Type in text box, press Enter; Requirements: None
Speech — How It Works: Click mic button, speak, click stop; Requirements: Microphone permission
Video — How It Works: Click video button, record, preview, submit; Requirements: Camera + mic permission
Image — How It Works: Click image button, take photo or select file; Requirements: None (camera optional)
AI asks follow-up questions based on responses
AI references earlier answers when relevant
No time limits on responses
Progress advances as the participant works through the questions
Can pause:
Close browser at any time
Progress saved automatically
Click link again to resume
See "Welcome back!" message
Limitations:
Must use same device
Must use same browser
Cannot switch devices
Clearing cache deletes session
If the connection is interrupted:
The participant is notified and can resend the message
Saved progress is preserved
Completion message displays (default: "Completed 🥳")
Input area disappears
Progress bar shows 100%
Conversation remains visible
Simple Thank You:
Shows completion message
No automatic redirect
Participant closes browser when ready
Panel Redirect:
Shows brief message
Automatically redirects to panel URL
Passes completion data
Triggers rewards/credits
Custom Next Steps:
Shows message with embedded links
May include follow-up surveys, calendar links, or incentive forms
If participant clicks link again:
Shows completion screen
Cannot add new responses
Can view conversation
Message indicates already participated
Answers are saved to Tellet as you give them. Stored locally in your browser (which is why resuming works on the same device only):
Session ID
Progress state
Language preference
Requirements:
Same device to resume
Same browser to resume
Session lost if cache cleared
Incognito sessions don't persist after closing
Responses save automatically as typed/recorded
No manual save action required
If the connection drops, your answer isn't lost — you can resend it when you're back online.
Supported:
A current version of Chrome, Firefox, Safari, or Edge
Not supported:
Internet Explorer
Very old or stripped-down browsers
Text — Best For: Quick answers, public spaces; Permission Needed: None; Upload Speed: Instant
Speech — Best For: Detailed stories, tone; Permission Needed: Microphone; Upload Speed: Fast
Video — Best For: Demonstrations, context; Permission Needed: Camera + Mic; Upload Speed: Slower (Wi-Fi recommended)
Image — Best For: Visual examples; Permission Needed: None; Upload Speed: Fast
Switching devices — not supported. The session marker lives in your browser, so the interview must be finished on the device where it started.
Returning after an interruption — on the same device and browser, click the link again to see a "Welcome back!" notification with all previous responses, then continue from the last question.
Network disconnection — handled gracefully: the participant is notified, can resend the message, and saved progress is preserved.
Permission denied — if the camera or microphone is blocked, clear browser-specific instructions help re-enable it, or the participant can continue with text-only responses.
Red "Test chat" banner visible
Restart button available
Responses are marked as test data and kept separate from real results
No quota consumption
Clean interface
No test indicators
Data permanently stored
Counts toward quotas