> ## Documentation Index
> Fetch the complete documentation index at: https://docs.qwairy.co/llms.txt
> Use this file to discover all available pages before exploring further.

# Is the AI perception of my brand aligned with our actual positioning?

> Compare themes in selected stored answers with your approved positioning, then investigate gaps without treating model output as market truth.

## Direct answer

Use **Analyze > Brand Perception** and the underlying stored answers to compare observed brand attributes with a current, approved positioning statement. Record alignment and divergence by prompt, provider, topic, and period. Treat the result as a view of the monitored dataset, not a complete measure of market perception.

## Data required

You need approved positioning claims with an owner and review date, a stable configured-prompt set, selected providers and models, stored answers, topics or tags, period, sentiment or perception outputs, and recorded citation URLs.

## Workflow

<Steps>
  <Step title="Define the reference">
    List the positioning attributes you intend to communicate and the evidence that supports each one. Exclude aspirational claims that are not yet true.
  </Step>

  <Step title="Select comparable answers">
    Choose prompts, providers, topics, and a period relevant to those attributes. Keep the cohort stable for later comparisons.
  </Step>

  <Step title="Review perception and evidence">
    Use Brand Perception to identify recurring observed themes, then read representative answers and their citations before classifying a gap.
  </Step>

  <Step title="Document the gap">
    For each attribute, record aligned, missing, contradictory, or unsupported output with the answer and filter context.
  </Step>
</Steps>

## Interpretation

A recurring attribute describes stored answers in the selected cohort. It does not establish what every user, provider, or market participant believes. Provider differences may reflect prompts, models, web access, sources, or sampling; the data alone does not identify the cause.

## Possible next actions

* If an intended attribute is missing, test whether an accurate first-party evidence page would make the positioning easier to verify.
* If an unsupported attribute recurs, investigate the cited sources and your own public copy before choosing a correction.
* If divergence appears in one provider slice, repeat the same prompt cohort before changing the broader positioning strategy.

## Limitations

Brand Perception is an analysis of observed outputs, not an independent reputation survey. Model outputs and source sets can change. The MCP surface follows its own documented tools and schemas; do not assume it returns the same aggregation as the product.

## Related pages

<CardGroup cols={2}>
  <Card title="Brand Perception" href="/documentation/analyze/brand-perception">
    Review the product analysis and its scope.
  </Card>

  <Card title="Answer Details" href="/looker-studio/sources/answer-details">
    Report on stored answers with documented dimensions.
  </Card>

  <Card title="Qwairy MCP" href="/mcp/introduction">
    Use supported MCP tools and schemas.
  </Card>

  <Card title="What AI says" href="/use-cases/reputation/what-does-ai-say-about-my-brand">
    Read the underlying stored answers.
  </Card>

  <Card title="Choose what to create" href="/use-cases/content/what-content-to-create">
    Form a content hypothesis from a verified gap.
  </Card>

  <Card title="Strengths and weaknesses" href="/use-cases/reputation/strengths-and-weaknesses-in-ai">
    Group recurring observed attributes.
  </Card>

  <Card title="Negative sentiment by topic" href="/use-cases/reputation/negative-sentiment-on-topics">
    Inspect topic-specific sentiment separately.
  </Card>
</CardGroup>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.