> ## 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.

# Which topic + provider combination is my biggest competitive weakness?

> Use GEO Matrix with one defined metric and stable filters to identify topic-provider cells that warrant answer-level review.

## Direct answer

Open **Cockpit > GEO Matrix**, select the metric and Topics as the row grouping, and review provider columns under a stable scope. Inspect unexpected cells and their underlying answers before calling any combination a weakness.

## Data required

Use the selected product metric with its exact formula, topics, configured prompts, providers, models where applicable, period, active list and filters, SELF/DIRECT classifications, stored answers, and citation URLs.

## Workflow

<Steps>
  <Step title="Choose one metric">
    Select the metric that matches the decision. Do not compare cells whose metric, denominator, or filters differ.
  </Step>

  <Step title="Review the matrix">
    Scan topic rows and provider columns for relative differences. Treat an empty cell as missing usable data until checked.
  </Step>

  <Step title="Inspect candidate cells">
    Open the cell detail and verify answered prompts, competitors, sources, and underlying stored answers.
  </Step>

  <Step title="Prioritize with context">
    Combine the observed gap with business relevance, cohort size, evidence quality, and feasible ownership.
  </Step>
</Steps>

## Interpretation

Each cell applies the selected metric to one topic and provider intersection under active filters. Color is comparative context, not a universal performance threshold. A low cell can reflect limited answered prompts, a changed provider mix, or an actual SELF/DIRECT difference.

## Possible next actions

* If a cell has sparse data, test the monitoring cohort before planning remediation.
* If DIRECT occurrences dominate a well-supported cell, inspect answers and sources to form a competitive hypothesis.
* If an entire topic differs across providers, test whether the prompts and model coverage are comparable.

## Limitations

The matrix represents configured monitoring data, not every topic or provider output. Provider and model changes can affect cells. The REST tags endpoint is a separate contract and does not by itself reproduce the GEO Matrix or its product metric.

## Related pages

<CardGroup cols={2}>
  <Card title="Tag Performance" href="/looker-studio/sources/tag-performance">
    Build a related report from documented tag data.
  </Card>

  <Card title="Tags API" href="/developers/endpoints/tags/list">
    Read the separate REST tags contract.
  </Card>

  <Card title="Shared Links" href="/documentation/workspace/shared-links">
    Share the configured matrix view.
  </Card>

  <Card title="Highest-priority actions" href="/use-cases/reporting/highest-priority-actions">
    Turn verified observations into a work queue.
  </Card>

  <Card title="Brand presence by topic" href="/use-cases/visibility/for-which-topics-does-my-brand-appear">
    Analyze topic-level SELF presence.
  </Card>

  <Card title="Page Performance score" href="/use-cases/reporting/unified-ai-score-per-page">
    Review page-level signals separately.
  </Card>
</CardGroup>


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