🔭

Quick competitive scan

Go from a stack of competitor pages to a clean comparison table in one pass — cheaply.

This is a high-volume, well-defined extraction task — the textbook case for a fast, cheap tier rather than a flagship. The win is a tight format constraint so the output stays scannable and short.

Gemini

Gemini 3.7 Flash

Why this variant: Fast and able to take all the source pages at once, so you skip summarising each. The tight cell limit keeps the table usable and the output cheap.

Prompt for Gemini 3.7 Flash

{paste 5–6 competitor pages or notes, separated by ---}
Build a comparison table with columns: competitor | positioning |
pricing model | one clear differentiator. Keep every cell under 12 words.
No commentary before or after the table.

Claude

Claude Haiku 4.5

Why this variant: When you're running this repeatedly across many competitors, Haiku does the extraction at a fraction of the cost. Give it a rubric, not an open-ended question.

Prompt for Claude Haiku 4.5

Extract these fields for each competitor below:
name | positioning (≤10 words) | pricing model | primary audience.
Output one row per competitor, pipe-separated. No extra text.
<competitors>
{paste competitor descriptions}
</competitors>