You’re in good company
CustomerSure uses AI to analyse and summarise customer feedback, not to make decisions or take actions on your behalf.
This page explains, in practical terms:
If you are new to core reporting concepts such as topics, labels, and comment boxes, you may want to read Reporting basics first.
AI in CustomerSure is used to interpret and condense free-text customer feedback.
Specifically, it can:
AI output always appears alongside the original customer comments. The comments remain the source of truth.
AI in CustomerSure does not:
AI analysis and summaries are advisory. They help you get more work done more quickly, support insight, but they do not determine outcomes.
AI is currently applied in four places:
If AI is enabled on a comment box, CustomerSure will analyse the text customers enter and:
You can enable or disable AI per comment box. Rating questions, choice questions, and text fields are not analysed.
See Adding and editing survey questions for details.
Topics represent things your organisation does that customers may feel positively or negatively about.
When analysing comments, AI uses your topic structure and wording to decide:
See Topics and sub-topics for guidance on structuring topics.
Labels categorise feedback without sentiment (for example: Complaint, Needs reply, Legal risk).
AI can apply both built-in and custom labels by interpreting:
Labels can always be added or removed manually.
See Labels for setup details.
AI can generate summaries of groups of feedback, for example:
Summaries are created by analysing the underlying customer comments and describing the most common themes and sentiments found.
Summaries do not replace reading feedback. They are a great tool to quickly engage stakeholders and help operational teams make sense of huge data sets, but reading customer comments is what gives you the deep insight you need to make changes in the business that improve the customer experience.
Modern language models are very good at analysing customer feedback. AI classification and summarisation is accurate the majority of the time.
However:
For this reason, AI output should be treated as strong guidance, not unquestionable fact.
AI behaviour in CustomerSure is strongly shaped by how you configure the system.
Clear, specific topic names lead to clear, specific analysis.
Good topic name
Poor topic name
If a topic name is vague to a human, it will be vague to the AI.
By grouping sub-topics into top-level topics, you can make your intent even clearer to the AI analysis engine.
Once your topics have clear names and a logical structure, make sure you provide clear descriptions for each topic — see the topics guide for more on how to do this.
For custom labels, the description is also critical.
AI uses the description to understand when the label should apply.
Good description
Apply when the customer explicitly requests follow-up or asks a question that requires a response.
Poor description
Important feedback.
CustomerSure is designed so AI analysis and summaries can always be reviewed.
Recommended practices:
If summaries or classifications feel misleading, ask us for help structuring your topic list or writing good topic descriptions.
Getting the most value from AI depends on getting the fundamentals right: topic structure, clear labels, and sensible survey design.
If you would like help reviewing or improving how AI behaves in your account, please contact our support team. We’re happy to help.
