I’m trying to figure out which prompts we should create content for because they trigger competitor mentions. How do I find them?

A good way to find those prompts is to look for queries that surface competitor names in the output or in the sources the model cites, then group them by intent.

Here’s a practical workflow:

1) Start with your competitor list

Make a list of:

• Direct competitors
• Adjacent alternatives
• Common category names users might use instead of brand names

2) Collect your prompt logs

Pull prompts from:

• Chat history / support logs
• Search logs
• On-site AI usage logs
• Sales/demo transcripts
• SEO keyword tools or FAQ queries

3) Run a competitor-mention scan

Search prompts for:

• Competitor brand names
• Product names
• Misspellings / abbreviations
• “best X” / “compare X” / “X vs Y” patterns

Also check model outputs, because sometimes the prompt itself doesn’t mention a competitor, but the response does.

4) Cluster by intent

Common buckets:

• Comparison: “X vs Y”
• Alternatives: “best alternatives to X”
• Migration: “switch from X to Y”
• Evaluation: “is X better than Y”
• Problem-solving: “how to do [task] in X”

These are often the highest-value prompts to create content for.

5) Prioritize by volume and business value

Score each prompt cluster by:

• Search volume / frequency
• Conversion potential
• Stage in funnel
• How often competitor names appear
• Ease of creating useful content

6) Use “trigger” reports

If you have analytics tooling, create a report that flags:

• Prompt contains competitor name
• Prompt leads to competitor mention in response
• Prompt is in a high-intent category

That gives you a queue of prompts to target with content.

7) Validate manually

For the top clusters, review a sample of prompts and outputs to make sure:

• The competitor mention is real and meaningful
• The intent is worth targeting
• The content opportunity isn’t already covered

If you want, I can help you build:

• a keyword/query classification template
• a spreadsheet schema
• or a SQL/regex approach to identify competitor-triggering prompts from logs.