campaign-strategy

Provide the client's website URL and optionally a client-profile.yaml with ICP and offer details. The skill crawls the site, analyzes case studies, and outputs 15-25 campaign ideas with targeting filters, AI personalization strategies, and value propositions in a markdown table.

Use publicly available data: website content, LinkedIn activity and engagement, job postings, technology stacks via BuiltWith or Wappalyzer, news, funding announcements, and Google reviews. Avoid scraping G2, Capterra, or Trustpilot reviews, which are not reliably accessible.

Sourcing pulls a new list where every record matches a hard criterion, such as companies that raised Series A. Enriching takes an existing list and adds data like funding or tech stack to filter and personalize, which is often cheaper and more scalable.

Only invert signals that are reliably detectable, such as job titles or job postings. Absence of a detected technology is unreliable, since failing to detect a tool does not prove the prospect does not use it.

It saves a campaign-strategy.md file to profiles/<business-slug>/ containing a customer discovery analysis, a campaign ideas table ordered broadest to most niche, no-AI campaign options, and front-end offer suggestions ready for copywriting handoff.