Search intent is easy to misread. These three prompts take raw keyword data and turn it into a gap map, a rewritten meta pass, and a topic cluster outline. Pick the one that matches where you are in the week.
[Claude]
I run a SaaS blog and exported 40 keywords ranked 10-20 from Google Search Console. You are my SEO lead. Group them by search intent using this scale: informational, commercial investigation, transactional, navigational. For each group, flag the pages that currently target the wrong intent, and list the three highest-leverage gaps where intent and page type clash. Output as a markdown table with columns: keyword, current intent, correct intent, misfit pages. Keep it under 60 rows and name the single gap I should fix first.
[ChatGPT]
Here are 25 page titles and their current meta descriptions from an ecommerce category. Rewrite each one to signal purchase intent while staying under 155 characters and 60 characters respectively. Preserve my brand voice: short, confident, no hype words like ultimate or best. Group the output by category, and add a one-line note where the title and description repeat the same keyword so I can diversify. Deliver a copy-paste table with columns: page, new title, new meta, reason for change.
[Cursor]
In this repo, I have a markdown file with 30 loosely related blog post ideas. Detect thematic clusters using k-means on simple keyword overlap, then for the strongest cluster write a pillar page outline with a main heading, 5 subheadings, and 3 internal links to existing posts. Save the outline to docs/pillar-outline.md and log the cluster sizes. Return a short summary of which clusters were strongest and which ideas are orphans.
Want every pack? Grab the archive at aitoolalliance.com/tag/prompt-pack.