The Guest Sentiment Watchlist agent uses AI to flag guest conversations that show signs of frustration during or shortly after a stay. This helps you proactively address issues and step in before a negative experience turns into a bad guest review.
The agent reads recent messages, rates the conversation's sentiment, and delivers a ranked report highlighting which threads need attention first. It only flags conversations for review and doesn't reply to guests or change a conversation thread.
Limitations
The agent analyzes conversations for reservations starting within the next 48 hours, currently underway, or recently ended. It doesn't evaluate conversations outside this window.
Understand the Guest Sentiment Watchlist agent
The Guest Sentiment Watchlist agent runs automatically once a day at 07:00. It can't be run manually.
During each run, the agent does the following:
- Reads recent messages for stays starting in the next 48 hours, underway, or recently ended.
- Rates conversations based on signals like unresolved complaints, repeated requests, and threats to cancel or leave a bad review.
- Highlights the conversations that need attention first and flags unanswered threads.
- Delivers a ranked report with direct links to each conversation in the Guesty Inbox.
Review the activity feed
You can view the results of the agent's daily run in the Agents hub activity feed.
Step by step:
- Sign in to your Guesty account.
- In the side navigation menu, select
Agents hub. - In the left submenu, click Agent activity.
- Click the relevant row with the agent's name to review the report details in the right-side panel.
Understand the report status and details
The report will have one of the following statuses:
- Done: No conversations showed signs of guest frustration, and no action is required.
- Action needed: One or more conversations showed signs of guest frustration and require your review.
The agent's report contains the following information:
- A summary of how many conversations were flagged and whether any were deemed critical and left unanswered.
- The number of severe conversations identified and whether they have received a reply, along with the total number of conversations scanned.
- A list of flagged conversations with details including the conversation, property, severity status, stay status, date, assignee, and a summary of why the conversation was flagged.