What it is.
It reads survey responses from Google Sheets, groups them into positive, neutral, and negative, then batches them for analysis so the model reads a set instead of one answer at a time.
An AI step generates themes and insights per group, results aggregate into a consolidated report, and that report goes to a summary sheet and a Slack channel on a daily schedule.
What you get.
- Sentiment grouping across every response before any analysis runs.
- Batch-based analysis, so themes come from a set of answers.
- Themes and insights per sentiment group.
- A consolidated report written to a summary sheet and posted to Slack daily.
How to set it up.
Point the sheet nodes at your own survey data and replace the placeholder sheet ID and name.
Rename the expected columns, since the template arrives with non-English column headers you have to change.
Check the batch size, because too small a batch produces themes from noise.
Add your model credentials and run one day of responses before scheduling it.
Set the Slack destination to a channel where somebody owns the follow-up.
Keep the raw responses, since a theme is a summary and the quote is the evidence.
Use cases
Themes from a set, not a row
Batch the responses so the analysis sees them together.
Group sentiment first
Sort the responses before any theme work runs.
Report to the channel
Send the summary to Slack so the team reads it without a doc.
Best for
Open-text survey questions
Free-text answers tell you the most and get read the least.
Continuous feedback programs
A daily run turns a rolling survey into a rolling read on it.
Reporting themes upward
Three themes per sentiment group is a report format that survives a leadership meeting.