Skip to main content

​
Direct answer

Open Monitor > Response Analysis, filter to your brand and intended scope, and read the stored provider outputs. Summarize recurring claims only after checking the configured prompts, provider and model, topic, date, sentiment, and cited URLs.

​
Data required

Use configured prompts, stored answers, provider and model identifiers, run dates, topics and tags, SELF/DIRECT mention classification, sentiment or perception outputs, and citation URLs. Define the review period before exporting or sharing.

​
Workflow

1

Set the review scope

Choose the business question, prompt set, providers, models, topics, and period. Avoid combining unrelated cohorts.
2

Read representative answers

Review answers with SELF mentions as well as relevant answers where only DIRECT brands appear. Keep mentions and citations separate.
3

Code recurring themes

Group supported claims, opinions, omissions, and sentiment patterns. Record how many answers support each theme and retain examples.
4

Verify material claims

Compare factual statements with approved sources and flag anything outdated, ambiguous, or unsupported for follow-up.

​
Interpretation

A stored answer is one observed provider output for a configured prompt and run. It is not a statement from all users or all models. Product Mention Rate is answers with a SELF mention divided by answers with at least one SELF or DIRECT brand mention; it does not measure sentiment or citation presence.

​
Possible next actions

  • If a recurring claim is accurate and relevant, test whether it aligns with intended positioning.
  • If a factual claim is wrong, follow a documented correction workflow using authoritative evidence.
  • If an important topic has few answers, review the monitoring design before concluding that no narrative exists.

​
Limitations

Outputs can vary between runs and model versions. Automated themes and sentiment require review against the original answer. Qwairy observes the selected stored dataset and cannot make exhaustive claims about what AI says everywhere.

Exports

Export the selected answer cohort.

Answer Details

Create reporting from stored answer records.

Shared Links

Share a filtered product view.

Perception versus positioning

Compare observed themes with approved positioning.

Strengths and weaknesses

Group recurring observed attributes.

Provider consistency

Compare equivalent cohorts across providers and models.

Topics with brand presence

Analyze SELF presence by topic.