gaps

No clear leader

The prompts, topics and segments where AI engines don't recommend one brand most of the time: open ground the brand can take first.

When to use

• "Which use cases have no clear winner yet?", "Where is nobody leading?", "Unmet needs."
• To work on one prompt a competitor already wins, the Beat a competitor skill fits better.

What to read

1. list_topics and list_prompts: the brand's topics and tracked prompts.
2. get_prompt_performance: the brand's place on each prompt, per engine.
3. list_ai_answers for the prompts: which brands each engine recommends.
4. get_competitor_comparison: where competitors lead, to tell an open prompt from a taken one.
5. get_prompt_volumes: how often each prompt is searched, where known.

Steps

1. For each prompt, count which brands the engines recommend across its answers.
2. A prompt has no clear leader when no brand is recommended in most of its answers. Group these prompts by topic or segment.
3. Order them by searches a month where known, then by how close the brand already is.
4. For each, suggest one action: a page to write or update, a question to answer, a source to be listed on.
5. Offer to add the top ones to Action Center with create_action; the app asks the person first.

Output

1. A table: Prompt or segment, Engines, Brands that appear, The brand's place, Searches a month, Suggested action.
2. The three to start with, and why.

Don't

• Don't call a prompt won or lost from one answer: look across its answers and engines.
• Never make up a volume: write "unknown" where none is known.
• Never promise the brand will lead after an action.