Skip to main content

​
Direct answer

Treat a brand as a candidate emerging competitor only when its observed presence increases across comparable prompts and periods. Verify aliases and answer context before adding or changing its competitor relationship. One mention is not a trend.

​
Data required

  • Competitor mentions from matched current and prior windows
  • A fixed prompt, provider, model, country, language, topic, and funnel scope
  • Mention occurrences, distinct answered prompts, and answer counts
  • Known aliases and current competitor relationships
  • Full answers where the candidate appears

​
Workflow

  1. In Monitor > Competitor Mentions, identify brands newly present or increasing in the selected scope.
  2. Compare equal windows using the same prompt cohort and monitoring configuration.
  3. Check whether the change appears across distinct prompts, topics, and providers rather than repeated mentions in one answer.
  4. Read the matching answers in Monitor > Response Analysis to confirm that the entity is a relevant alternative.
  5. Review aliases, subsidiaries, and ambiguous names.
  6. If appropriate, classify the brand for future comparison and mark the date of that configuration change.

​
Interpretation

Use both occurrence-level volume and prompt-level breadth. Repeated occurrences in one answer do not show broad emergence. A newly classified DIRECT competitor changes the eligible denominator for Product Mention Rate and Share of Voice, so trend continuity must be annotated. Describe the result as an observed increase in the monitored dataset, not market entry or market-share growth.

​
Possible next actions

  • Add a candidate to a watchlist or DIRECT comparison as a reversible monitoring hypothesis.
  • Test whether the pattern persists in another matched period.
  • Investigate the topics and criteria where the candidate appears.
  • Correct entity aliases before changing competitive strategy.

​
Limitations

  • Prompt or monitoring changes can create artificial emergence.
  • Provider output can vary between runs.
  • Entity detection can merge or split similarly named brands.
  • Qwairy observations do not establish company growth, revenue, or market share.