ad-to-pagelisted
Install: claude install-skill seercraft-io/seer-skills
# Ad to page
A brand that builds a page for one ad is telling you that ad matters. Done means:
`seer/reports/ad-to-page-<date>.md` mapping each rival's top ideas to their pages, with the
page types they use and the message-match gaps worth copying or avoiding.
Read `${CLAUDE_SKILL_DIR}/../seer-start/reference/house-rules.md` first. `$S` = `${CLAUDE_SKILL_DIR}/../seer-start/scripts`.
## Steps
1. `python3 $S/ideas.py seer/ads/<rivals>-<date>.json` → "Landing pages" per brand.
2. For each rival take the 5 destinations with the most live ideas, plus any URL with a tell in it
(`/pages/`, `lp`, `listicle`, `advertorial`, `vs`, a creator's name, a campaign code).
3. Fetch each page. Many stores block plain fetches (403): if so, use a browser tool if you have one; otherwise list the URLs and ask the person to paste the first screen's text or a screenshot. Never guess a page's content. Record: page type (product page, collection, custom landing page, listicle,
advertorial, quiz, comparison), headline, first-screen offer, proof shown, and the main CTA.
4. For each of the top ideas, compare ad → page: same promise? same offer? same visual? Score
match as strong / partial / broken, with one line why.
5. Note the patterns: how many ideas go to custom pages vs product pages, per brand.
## Output
- Table per rival: idea (cited) · destination · page type · page headline · match.
- **What they build pages for**: the campaigns that got their own page (often the ones they're push