Notion is a brilliant first CRM for bootstrapped teams—free, flexible, visible. But the moment you hit 300 contacts or need real pipeline forecasting, the gaps become expensive. The export is where it hurts most. You'll get a CSV that looks complete, then watch fields collapse mid-migration: relationship chains break, contact histories flatten, deal notes orphan into unlinked text, custom fields reference IDs that don't map anywhere. Teams moving to Pipedrive, Orin , or QuickBooks often discover this at day three of integration, when half their context is already lost. The good news: these breaks are predictable. We've mapped the nine field types that fail most consistently and built a clean migration path for each. The nine fields that orphan on Notion CRM export Not all Notion properties export equally. Some flatten completely, others create orphaned records, and a few introduce silent data corruption that you won't catch until you run your first report. 1. Relation fields (the contact-to-account parent) Notion relations are powerful—link a contact to their account, a deal to the stakeholder. On export, Notion delivers these as raw ID strings or text labels with no reference context. What you export: contact_123 or Acme Corp What the target expects: A matching account ID in the destination system What happens: The receiving platform sees unlinked text. Your contact floats orphaned. Safe mapping: Before export, create a lookup table. In Notion, add a formula field that outputs [Account Name] + "|" + [Account ID from your next system] . Export that combined string, then split it on import. Most platforms (Pipedrive, Orin ) have import mapping steps where you can match by account name first, then link on arrival. 2. Rollup fields (aggregated deal values, contact counts) Rollups in Notion calculate sums, counts, and concatenations from related records. They're live in Notion; they're static text on export. What you export: A frozen number like $450,000 or Contact Count: 12 What it was: A live formula that recalculated whenever a linked deal closed What happens: You import a static number. A week later, a new contact joins the account, and the number is now wrong. No one notices until reconciliation. Safe mapping: Don't export rollups directly. Instead, export the source data —the individual deals or contacts that feed the rollup. Let your destination system recalculate. If you must carry a summary field (like Total ARR ), export it as Imported ARR Snapshot - Do Not Edit and rebuild the live formula in the target platform. Pipedrive and Orin both support linked deal summaries; use those instead of static numbers. 3. Button fields (automations that don't exist in the target) Notion buttons trigger automations or webhooks. On export, they simply vanish—they have no value, no reference, no trace. What you lose: Any workflow that depended on Click to create quote or Log call What happens: Your team lands in Pipedrive with no shortcuts. Productivity dips 10-15% for two weeks while they relearn how to trigger actions. Safe mapping: Before you migrate, audit every button in your Notion CRM. Map each one to an equivalent action in your target. If Notion had a Create Zapier task button, Orin has native automations ; Pipedrive has workflows. Build these before import. Train your team on the new workflow. Don't leave buttons orphaned—replace them first. 4. Formula fields with Notion functions that don't map Notion formulas are powerful: dateBetween() , parseJson() , text manipulation. Most don't exist in other platforms. What you export: A calculated string or number ( Days to close: 45 ) What happens: The static result imports fine. But if you need to recalculate later (new year, new fiscal calendar), you have to rebuild the formula from scratch in the target, or maintain it in a spreadsheet on the side. Safe mapping: Export formula results only if they're snapshots you'll never need to change. For living calculations—contract value, days until renewal, pipeline stage probability—export the source fields and rebuild the formula in your destination. Most CRMs support formula fields or calculated properties. It's faster to rebuild once than to manage divergent logic across systems. 5. Lookup fields (displaying linked record properties) Lookups pull a property from a related record—for example, Account Phone showing the phone of the linked account. On export, you get the value, but the link is lost. What you export: (555) 234-5678 What you lose: The fact that this phone came from the account record. If the account phone changes, your contact record will show stale data. Safe mapping: Lookups can be re-created if you import the relation first. Export the lookup value separately, label it as Account Phone (from Account record) , and import it as a text field. Then rebuild the lookup in your target platform—link the contact to the account, then use a lookup or formula to pull the phone. This preserves the relationship and kee