An address scraped from a page is easy to paste into a spreadsheet. The hard part is answering the questions that follow: which site did it come from, which page published it, and what should happen when no address was found? A useful contact-research pipeline needs to preserve those answers as...
Source: [Dev.to](https://dev.to/nicedayup/a-website-list-is-not-a-contact-list-designing-a-reviewable-extraction-pipeline-4g6b)