{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/320038e70c3b4ba29ad5265a3e83fef3\" frameborder=\"0\" width=\"1668\" height=\"1251\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1251,"width":1668,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1251,"thumbnail_width":1668,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/320038e70c3b4ba29ad5265a3e83fef3-f061b349010db6cb.gif","duration":282.219,"title":"Building and Testing Lead Scrapers Pipeline","description":"This Loom explains the author’s lead-scraping and data-cleaning approach, focusing on why Clutch became the primary source. The repo shows lists totaling 100 leads, with seventh from Clutch and 29 from Current Base, after the author initially measured the first 300 Current Base companies and found 74 percent raised over 20 million and 77 percent matched the ICP. For Clutch, they built TypeScript and Playwright scrapers that iterate three Clutch listings, wait about two seconds between requests, stop on captcha, and check rate limits while writing results, addressing earlier issues like 300 empty rows. The output CSV includes per-row source and origin plus fields like email and signals, and a rule for interpreting “VP of sales” versus “wants sales” for agencies, with a pipeline that scores leads and pauses for human approval."}