{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/cb0520e936734a368741cf68c1dcfcf6\" frameborder=\"0\" width=\"1620\" height=\"1215\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1215,"width":1620,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1215,"thumbnail_width":1620,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/cb0520e936734a368741cf68c1dcfcf6-0b9bf3ea4ec7f1a5.gif","duration":386.255,"title":"Backfilling HubSpot Data via Claude Code","description":"This Loom explains how the team backfilled and remapped HubSpot channel subchannel values using a custom, multi-step script due to complex logic changes requested by the client. It updates prior “subchannel of Google” values such as PaidSearchGoogle, PaidDisplayGoogle, and OrganicSearchGoogle to new subchannel formats, while handling unmapped values like cases where latest subchannels equal Google but actually indicate email or referral by flagging them for architect review. The script ran a dry run, then performed batch updates after confirmation: 1,587 contacts and over 3,000 field updates across 16 successful batches, with minor HubSpot lag. It also produced CSVs and an MD summary documenting portal IDs, mapping approach, and results, and flagged a downstream workflow issue that required updates to the channel workflow."}