playbook-social-posts

Provide the company page or personal profile URL and run the async 3-call scrape pattern: start the run, poll for completion, then read the dataset. The playbook returns the post text, URL, days ago, and author type, with an optional one-line personalization on request.

Batch up to 100 profile URLs per run start, since each run start bills separately and a batch of 8 costs the same start fee as a batch of 1. Dedupe contacts to unique profiles before scraping and fan results back out, cutting cost from about $2.95 to $1.83 per 1,000 contacts.

Yes, one scraper accepts both URL types. Company posts fan out to every contact at that company, while person posts never fan out. Group results on universalName or publicIdentifier, because the company-style identifier is null on person items.

On person items the company-style identifier is null and the slug lives in publicIdentifier instead. Group on universalName or publicIdentifier and normalize URLs before joining, otherwise every person row silently vanishes with no error.

Testing showed 71% coverage on high-frequency B2B SaaS company pages, which is an upper bound. SMB, trades, and long-tail lists typically yield 20 to 40%, so measure fill on a 100-row sample of your own list before writing copy around it.

Spintax picks at random and cannot see an empty variable, so it renders broken text like "Noticed ." and discards personalization you already paid for. Split the list on a non-empty social_post_line before upload and rewrite the clause for the empty segment.