Interview Completion Flow
What happens when an interview ends — thank-you screen, participant feedback, and behind-the-scenes AI analysis.
When a participant finishes their interview, a completion flow kicks in that handles two things simultaneously: giving the participant a satisfying ending, and triggering Koji's analysis pipeline behind the scenes. Here is everything that happens from the moment the conversation wraps up.
How Interviews End
An interview can end in two ways:
Automatic Completion
The AI interviewer tracks which topics from your research brief have been covered and how thoroughly. When it determines that sufficient ground has been covered, it naturally wraps up the conversation — thanking the participant, summarizing the key themes discussed, and saying goodbye.
This is the most common path. The participant does not need to do anything; the conversation flows to a natural conclusion just like a real interview.
Manual Completion (the "Done" Button)
At any point during the interview, participants can click the Done button to end the conversation on their own terms. This is useful when:
- The participant feels they have said everything they want to say
- They need to leave unexpectedly
- They have reached the end of the topics they care about
Whichever method triggers the end, the next step is the same: the participant sees the thank-you screen.
The Thank-You Screen
Once the interview ends, participants see a clean, simple completion screen.
Visual Indicator
A green checkmark icon is displayed inside a green circle at the center of the screen, providing a clear visual signal that the interview is complete.
Thank-You Message
Below the checkmark, the heading reads "Thanks for the chat!" — a brief, warm message that acknowledges the participant's time without being overly formal.
The completion screen is intentionally minimal. There is no avatar, no personalization with the participant's name, and no display of interview duration. The focus is on a clean ending and a quick feedback step.
Participant Feedback
After the thank-you message, participants are invited to rate their experience with a simple binary choice:
Thumbs Up / Thumbs Down
Two buttons appear below the thank-you message:
- Thumbs up — labeled "Loved it"
- Thumbs down — labeled "Not great"
Participants tap one of the two options to share their sentiment. There is no numeric scale, no emoji range, and no written feedback text area — just a quick, low-friction binary choice.
Immediate Submission
Feedback is submitted the moment the participant taps their choice. There is no separate submit button or confirmation step. The participant taps "Loved it" or "Not great," the selection is recorded immediately, and they are done.
Feedback is not mandatory. If a participant closes the tab without tapping either button, the interview is still recorded as completed and the analysis still runs.
What Happens Behind the Scenes
While the participant is looking at the thank-you screen, Koji's analysis pipeline is already running in the background. Here is what happens automatically:
Transcript Processing
The full conversation — whether it was conducted via voice or text — is processed into a clean text transcript. For voice interviews, the audio is transcribed. For text interviews, the messages are already in text form.
Quality Scoring
Koji's quality evaluation assigns a score from 0 to 5 to the interview. This score reflects the depth, relevance, and informativeness of the participant's responses relative to your research brief.
- Score 3 and above: The interview counts toward your study and contributes meaningful data.
- Score below 3: The interview is flagged as low quality. It does not count toward your plan's interview limits, and you are not billed for it.
This quality gate ensures you only pay for interviews that actually deliver value. Learn more in How the Quality Gate Works.
AI-Generated Insights
After scoring, Koji analyzes the transcript to extract:
- Key themes discussed during the interview
- Notable quotes that capture important sentiments
- A summary of the participant's main points
- Sentiment indicators for different topics
- Structured question responses with their quantitative values (for text interviews that included structured question widgets)
These insights appear on the interview detail page in your dashboard and contribute to the project-level analysis.
Analysis Timing
The analysis typically completes within a few minutes of the interview ending. You will see the results appear in your project dashboard as they become available. Voice interviews may take slightly longer due to the additional transcription step.
What the Researcher Sees
Back in your project dashboard, completed interviews show up with:
- A status badge indicating the interview is complete
- The quality score (0-5)
- The participant's name (if collected)
- Message count and duration
- A link to the full transcript and analysis
You can click into any completed interview to read the transcript, review the quality score breakdown, and see the AI-generated insights.
Handling Abandoned Interviews
Sometimes participants start an interview but close the browser before finishing. Koji handles this gracefully:
- The conversation is saved up to the point where the participant left
- After a period of inactivity, the interview is marked as abandoned
- Abandoned interviews are still analyzed if they contain enough content to be meaningful
- They appear in your dashboard with an appropriate status indicator
Feedback Data for Researchers
The thumbs-up and thumbs-down feedback that participants submit on the completion screen is stored alongside the interview record. You can use this data to:
- Monitor participant satisfaction across your study
- Identify issues with specific interview topics or questions
- Improve future studies based on whether participants found the experience positive or negative
Feedback ratings are visible in the interview detail view and can be included in exports.
Customizing the Completion Experience
Currently, the thank-you screen uses a standard layout and messaging. The green checkmark, "Thanks for the chat!" heading, and binary feedback buttons are consistent across all studies.
If your interview is embedded via the embed widget, the completion screen renders inside the iframe. You can listen for the koji:interview_completed event to trigger your own follow-up actions on the parent page — such as showing a discount code, redirecting to another URL, or logging the completion in your analytics.
Next Steps
- How the Quality Gate Works — learn how interviews are scored and filtered
- Voice Interview Experience — understand the full voice interview flow
- Text Interview Experience — understand the full text interview flow with structured widgets
Further reading on the blog
- How to Conduct Remote User Interviews: The Complete Guide (2026) — Remote user interviews are now the default for most research teams. This guide covers everything — from recruiting and scheduling to running
Related Articles
Understanding Quality Scores
Learn how Koji evaluates interview quality on a 0-5 scale and why it matters for your research and billing.
How the Quality Gate Works
Understand Koji's quality gate — conversations scoring below 3/5 are completely free and don't consume credits, protecting your research budget.
How to Improve User Interview Completion Rates
Why user interview completion rates drop, what good benchmarks look like, and the 9 concrete levers that move the rate from 40% to 80%+. Includes the AI-moderation effect, modality choice, length tuning, and incentive design.
Voice Interview Experience
What participants see and hear during a voice interview — from microphone permission to natural conversation.
Text Interview Experience
How text-based interviews work for participants — chat interface, streaming responses, and conversation flow.
How to Customize Your Research Study: Branding, Landing Pages, and Intake Forms
Complete guide to Koji's study customization options — landing page design, branded colors, animated orbs, intake forms, trust badges, interview mode settings, and link previews.