Newsletter Sponsor & Positioning Brief

Analyze public newsletters for the compact sponsor, cadence, and positioning signals an AI agent needs before doing creator, sponsor, or market-research work.

This actor is built for public-only newsletter research. It does not log in, bypass paywalls, collect private subscriber data, read private archives, or use internal Napoleon's Corporal materials.

What It Returns

• Publication name and public URL
• Recent public post count
• Estimated posts per week
• Cadence bucket: regular, semi-regular, irregular, or insufficient data
• Last visible publication date
• Public paywall language ratio when visible in post summaries
• Visible subscriber or follower count when shown on the public page
• Topic keywords from public titles/descriptions
• Headline pattern signals
• Audience lane
• Editorial promise
• Sponsor fit summary
• Sponsor inventory summary
• Sponsor inventory status and confidence
• Sponsor signal count and density
• Sponsor language patterns
• Public sponsor mention examples
• Detected sponsor names when public text names them
• Commercial readiness score
• Cadence risk
• Paywall posture
• Evidence quality
• Brief confidence
• Recommended next action: pitch, qualify, monitor, benchmark, or skip
• Action rationale
• Best-fit sponsor categories
• Disqualifiers
• Next data needed
• Agent use case
• Batch rank, priority score, and priority tier when using batch qualification mode
• Plain-language positioning summary
• Field status object that marks unavailable data instead of guessing

Input

{
  "publicationUrls": ["https://www.platformer.news/"],
  "postLimit": 25,
  "analysisMode": "positioning_brief"
}

analysisMode can be:

• positioning_brief
• sponsor_fit
• sponsor_inventory
• batch_qualifier
• cadence_audit

Use sponsor_inventory when the buyer job is to find public sponsorship clues such as sponsor language, media-kit language, named sponsor mentions, and post-level sponsor snippets.

Use batch_qualifier when the buyer job is to pass a list of public newsletter URLs and receive a ranked prospect queue. Batch rows include batch_rank, batch_priority_score, batch_priority_tier, and batch_qualification_summary.

Output

Each publication produces one dataset item. The actor is designed for pay-per-result billing through the publication-result charged event.

The intended agent workflow is:

1. Run one public newsletter URL.
2. Read recommended_next_action.
3. Route the next agent step:
   • pitch: outreach or sponsor prospecting can start.
   • qualify: collect the missing data listed in next_data_needed.
   • monitor: keep it in a watchlist.
   • benchmark: use it as a comparison target, not a prospect.
   • skip: do not spend more agent calls on it.

For batch work, sort by batch_rank and start with rows marked top_priority or qualified.

Public-Only Boundary

This actor uses public HTML, public RSS/Atom feeds, and public metadata. Unavailable values are returned as null with a matching field_status explanation. It does not infer hidden engagement, private subscriber counts, paid-post content, email metrics, or account-only analytics.

Local Development

Install dependencies:

$python -m pip install -r requirements.txt

Run locally through Apify tooling or directly in an Apify-compatible environment. For deployment, use the Apify CLI from this actor directory after authenticating:

$apify push

Benchmark Before Publish

Run the local benchmark before any marketplace publication:

$python3 scripts/run_benchmark.py

The benchmark uses public newsletter pages and public Apify Store metadata. It does not run paid competitor actors.

Monetization Notes

Do not finalize pricing until benchmark output shows a clearer one-result brief than raw scraper plus generic summarization. If the benchmark passes, the candidate price is closer to $0.01 per publication brief than $0.002 per raw result. Do not claim revenue until a paid Apify run has been verified.