> ## 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.

# What are the strengths and weaknesses AI associates with my brand?

> Group recurring positive and negative attributes in selected stored answers, then verify their context and evidence before acting.

## Direct answer

Use **Analyze > Brand Perception** to identify recurring attributes, then read the answers that support each theme. Classify an attribute as an observed strength or weakness only for the selected prompts, providers, models, topics, and period.

## Data required

Use stored answers, configured prompts, provider and model, period, topics and tags, perception or sentiment outputs, SELF/DIRECT mentions, citation URLs, and your approved positioning and product facts.

## Workflow

<Steps>
  <Step title="Choose a representative cohort">
    Select prompts that cover the decisions and topics you care about. Avoid mixing unrelated audience needs.
  </Step>

  <Step title="Review recurring attributes">
    Use Brand Perception to locate repeated positive and negative themes. Retain answer counts and filters.
  </Step>

  <Step title="Read and verify">
    Inspect representative answers and citations. Separate supported facts, subjective trade-offs, outdated claims, and classification errors.
  </Step>

  <Step title="Compare with positioning">
    Map each verified theme to an approved positioning attribute and note aligned, missing, or conflicting observations.
  </Step>
</Steps>

## Interpretation

Frequency in stored answers indicates recurrence in the monitored dataset, not objective product quality or broad market opinion. Positive sentiment is not equivalent to a recommendation, and a negative trade-off may be appropriate for an audience outside your positioning.

## Possible next actions

* If a supported strength is absent from important prompts, test clearer evidence or positioning on relevant first-party pages.
* If a weakness is factual and outdated, test the misinformation correction workflow.
* If a weakness is a valid trade-off, test content that explains intended fit rather than trying to suppress it.

## Limitations

Perception and sentiment analyses can simplify nuanced language. Results depend on prompt design, provider and model mix, and the selected period. They do not measure customer experience or predict commercial outcomes.

## Related pages

<CardGroup cols={2}>
  <Card title="Brand Perception" href="/documentation/analyze/brand-perception">
    Review observed perception themes in the product.
  </Card>

  <Card title="Answer Details" href="/looker-studio/sources/answer-details">
    Analyze the supporting answers in Looker Studio.
  </Card>

  <Card title="Shared Links" href="/documentation/workspace/shared-links">
    Share a filtered review with stakeholders.
  </Card>

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

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

  <Card title="Perception versus positioning" href="/use-cases/reputation/ai-perception-vs-positioning">
    Compare themes with approved positioning.
  </Card>

  <Card title="Correct misinformation" href="/use-cases/reputation/correct-ai-misinformation">
    Address verified factual errors.
  </Card>
</CardGroup>


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