Skip to main content

Verbatim analytics

Verbatim analytics turn the free-text comments your customers leave into something you can read at a glance: an overall mood, the words people use most, and the topics they keep raising. Everything lives on the Verbatim page, organised into five tabs.

Pick a survey in the filter bar at the top of the page first — verbatim analysis is always scoped to a single survey.

The five tabs

TabWhat it showsWhen to use it
SentimentAn AI-scored positive / neutral / negative breakdown, plus a sentiment trendCheck the overall mood and how it is moving over time
Word cloudThe most frequent words, sized by how often they appear, with an excluded-words listSpot recurring language without reading every comment
TopicsCoded comments grouped by topic, under their parent themeSee which themes and topics dominate the feedback
CodingThe tagging workbench — apply topics to individual comments, or let the AI do itBuild and refine the structured view your Topics tab reports on
Review queueLow-confidence AI codings, worst first, ready to approve or rejectQuality-check the AI's work where it was least sure

How the pieces fit together

  1. Comments arrive with each survey response and are stored against the question that asked for them.
  2. The AI scores each comment's sentiment (see Sentiment).
  3. The AI also auto-codes comments against your topics and themes — the structured taxonomy you define in Admin (see Topics and themes).
  4. You review, correct, or add to that coding on the Coding tab, and quality-check the AI's least-confident tags on the Review queue tab.
  5. The Word cloud and Topics tabs visualise the result (see Word cloud and charts).
tip

Sentiment scoring and auto-coding happen automatically as responses are captured. The buttons on the page ("Analyse with AI", "Auto-tag with AI") are there to catch up any comments that have not yet been processed — for example, after a bulk import.

Anonymity

To protect respondent confidentiality, verbatim figures are suppressed when there are too few responses to report safely. When this happens you will see a note such as "Below the anonymity threshold" and the score shows as a dash () instead of a number. Gather more responses and the figures will appear.

Exporting

The Sentiment tab includes an Export coded CSV button, which downloads the verbatim comments together with their applied topic codes. This is handy for sharing raw feedback with colleagues who do not use the platform.

Next steps