Notion is outstanding for knowledge work. But as a CRM, it hits a wall at around 500 contacts—and that wall becomes a cliff the moment you try to move. Export your Notion database and watch what happens: every relational field (your "Contact belongs to Company" link, your "Deal is connected to Contact" reference) flattens into a text string or a comma-separated list. Your relationship graph doesn't migrate; it collapses into orphaned records and broken lookups. The difference between a good migration and a failed one isn't moving data—it's understanding what data actually travels, and what you have to rebuild by hand. Here's the map. Why Notion's export breaks relationships (and what actually happens) Notion's relational fields are its strength inside Notion. A "Contact" database linked to a "Company" database shows you instantly which people work where. But Notion exports to CSV—and CSV has no concept of relationships. When you export: One-to-many links become text. A contact linked to three deals shows up as "Deal A; Deal B; Deal C" in a single cell. No structure, no parent-child hierarchy. Multi-select and rollup fields lose their logic. Anything computed or cross-referenced collapses into a static snapshot. If you had a rollup showing "total deal value for this contact," that number won't update once it's in a spreadsheet. Timestamps and audit trails vanish. Notion's "created by," "last edited by," and modification history don't export. You lose the timeline of who did what. Custom field metadata gets lost. A field that was a number in Notion becomes text in the CSV. A date becomes a string. You have to re-type the field definitions. The practical outcome: you land in Orin with orphaned contacts, no company-contact relationships, and deals floating unattached to anyone. The data is there—the structure isn't. Step 1: Map your Notion databases to Orin's three-tier model (contacts → companies → deals) Orin's CRM is built around three primary objects: contacts , companies , and deals . Each has a parent-child relationship. A deal lives under a company. A contact lives under a company. A deal has a primary contact. This isn't arbitrary—it's how revenue actually flows. Your Notion export usually has at least two databases: one for people, one for organizations or opportunities. Map them like this: Notion "People" database → Orin Contacts Notion "Companies" or "Organization" database → Orin Companies Notion "Deals" or "Opportunities" database → Orin Deals If you have a fourth database (e.g., "Interactions," "Meetings," "Notes"), those map to Orin's activity log and timeline—which live inside each contact and deal record, not as separate objects. The key insight: do not import all three in parallel. Companies must import first. Contacts second (so they can link to companies). Deals last (so they can link to both contacts and companies). Order matters. Step 2: Export Notion, then map each field—and flag what's lost Export each Notion database as CSV. Open the file and create a mapping document. For every column in your Notion export, write down: The Notion field name The Orin field it maps to (or: "custom field in Orin") Whether it needs transformation (e.g., date format, phone number cleanup) Whether it's linked data that broke on export Here's what a realistic People database mapping looks like: Notion "People" field → Orin Contact field: First Name → First Name Last Name → Last Name Email → Email Phone → Phone Company (was a linked field) → Company ID (you'll populate this manually after import) Role → Custom field: "Job Title" LinkedIn → Custom field: "LinkedIn URL" Last Contacted → Activity log (you'll update this during import) Meeting Notes → Notes field (paste here) Recurring Revenue (rollup) → Custom field: "Est. Annual Value" (will be static) Active? (checkbox) → Contact Status: "Active" or "Inactive" The row marked "was a linked field"—that's the one that matters. In Notion, you had a relationship. In the CSV, you have a company name as text. You'll rebuild that link by hand. Step 3: The 14-day sequence—what recovers automatically, what needs manual work Days 1–2: Companies first. Import your Notion Companies database into Orin. Map the fields exactly. Anything with revenue, industry, or headcount goes into custom fields. Once companies are live in Orin, you have the parent records that everything else will hang from. Days 3–5: Contacts with company linkage. Import your Notion People database. For the "Company" column (which is now text, not a link), use Orin's bulk import feature to match by company name and auto-link. This works if your company names are clean and consistent. If they're not ("ACME Inc." vs. "Acme" vs. "Acme, Inc."), you'll do a manual reconciliation pass—count on 2–3 hours for 500 contacts. Days 6–9: Deal import and contact mapping. Import your Deals database. Again, the linked fields (contact name, company name) are text. Use bulk matching to link deals to companies and contacts by