Responses

Viewing Responses

Browse, search, and drill into individual interview responses.

Viewing Responses

The Responses page is where you review, analyze, and drill into individual interview responses. Each response includes an AI-generated summary, sentiment analysis, extracted keywords, notable quotes, and the full chat transcript.

Accessing Responses#

Navigate to the Responses page:

  1. From the sidebar: Select your project from the left navigation
  2. Click "Responses": This opens the response list for that project
  3. Filter or search: Use the date range, status, or sentiment filters to narrow results

All responses are displayed in reverse chronological order (newest first).

Response List: What You See#

Each response in the list is displayed as a card with key information at a glance:

Status Badge#

Indicates the current state of the interview:

  • Opened (Gray): The respondent loaded the page but hasn't sent a message yet
  • Started (Blue): The respondent sent at least one message but hasn't completed the interview
  • Completed (Green): The interview finished with a full AI-generated summary

Sentiment Badge#

Shows the overall sentiment for the response:

  • Positive (Green): The respondent expressed satisfaction or enthusiasm
  • Negative (Red): The respondent expressed frustration or criticism
  • Neutral (Gray): The respondent provided balanced or factual feedback
  • Mixed (Yellow): The respondent expressed both positive and negative sentiments

Sentiment is calculated by aggregating the sentiment of all answers in the conversation. Completed responses always have a sentiment label.

Summary Preview#

The first few lines of the AI-generated summary are displayed in the card. This gives you a quick snapshot of what the respondent said without opening the full response.

Example:

"The user appreciates the ease of onboarding but finds the pricing page confusing. They mentioned slow loading times and requested better documentation..."

Metadata#

Additional context displayed on each card:

  • Date and Time: When the interview was completed (or last updated if incomplete)
  • Respondent Email: If email collection is enabled and the respondent provided their email
  • Country Flag: The respondent's country based on IP geolocation

Opening a Response: Detail Drawer#

Click on any response card to open the detail drawer, which shows the full interview data.

Full AI-Generated Summary#

The top section displays the complete AI summary, which includes:

  • Overall Assessment: A high-level summary of the respondent's feedback
  • Per-Question Answers: Structured answers to each question in your interview
  • Key Themes: Recurring topics or patterns mentioned by the respondent

The summary is formatted for readability, with bullet points, sections, and clear headers.

The AI summary distills a 10-minute conversation into a 2-minute read. Use it to quickly understand the main points without reading the full transcript.

Sentiment and Confidence Per Answer#

Below each question's answer, you'll see:

  • Sentiment Label: Positive, negative, neutral, or mixed
  • Confidence Score: A percentage indicating how confident the AI is about the sentiment classification

Example:

Question: "What do you like most about our product?" Answer: "The onboarding process was incredibly smooth and intuitive." Sentiment: Positive (95% confidence)

Notable Quotes#

The AI highlights standout quotes from the conversation—phrases or sentences that capture the essence of the respondent's feedback.

Example:

"The UI is beautiful, but I got stuck on the pricing page and almost didn't sign up."

Quotes are displayed as pull-quotes for easy scanning and sharing.

Positive and Negative Keywords#

Keywords are automatically extracted from the conversation and categorized by sentiment:

  • Positive Keywords: Topics associated with satisfaction (e.g., "onboarding", "ease of use", "support")
  • Negative Keywords: Pain points and frustrations (e.g., "slow loading", "confusing UI", "pricing")

Keywords help you quickly identify what worked and what didn't without reading the full transcript.

Urgent Issues#

If the AI detects an urgent issue (e.g., a bug, security concern, or churn risk), it flags it at the top of the response with a red banner.

Example:

Urgent: The respondent mentioned a critical bug in the checkout flow that prevented them from completing their purchase.

Urgent issues can trigger automations to send Slack alerts, create tickets, or escalate to your team in real time.

Full Chat Transcript (Expandable)#

At the bottom of the detail drawer, you'll find the full conversation transcript:

  • Expandable Section: Click "Show full transcript" to see the entire conversation
  • Message-by-Message: Displays every question the AI asked and every answer the respondent gave
  • Timestamps: Each message includes a timestamp showing when it was sent

Use the transcript to:

  • Understand the context behind a specific answer
  • See how the AI adapted its follow-up questions
  • Identify edge cases or unexpected user behavior

Real-Time Updates#

The response list updates in real time as interviews are completed:

  • Auto-Refresh: New responses appear automatically without needing to refresh the page
  • Live Status Updates: Watch as responses move from "started" to "completed"
  • Notifications: If enabled, you'll receive browser notifications when new responses come in

Leave the Responses page open during a launch or campaign to monitor incoming feedback in real time.

What You Can Do With Responses#

Here's how teams use the Responses page:

Triage Feedback by Priority#

  • Filter by negative sentiment to identify urgent issues
  • Look for urgent issue flags to prioritize critical problems
  • Sort by date to review the most recent feedback first

Share Quotes With Your Team#

  • Copy notable quotes and share them in Slack, email, or product docs
  • Use quotes in presentations or reports to highlight user feedback

Drill Into Specific Themes#

  • Search for keywords like "pricing" or "onboarding" to see all responses mentioning that topic
  • Use the keyword analysis to identify recurring themes across multiple responses

Track Completion Rates#

  • Filter by "started but not completed" to understand where respondents drop off
  • Adjust your questions or probing intensity based on drop-off patterns

Next Steps#