Migrating from HubSpot

HubSpot’s contact/company/deal model maps cleanly onto Primicornis when you treat objects as canonical and pipelines as opportunity graphs.

Source → Primicornis mapping

HubSpot	Primicornis target
Contact	`person` (or custom)
Company	`company` or `startup` depending on thesis
Deal	`opportunity` with stage ↔ pipeline mapping
Owner	Map to workspace members via email before import
Associations	Use relation properties; maintain association labels as text metadata

Property mapping checklist

Export HubSpot properties with internal names (hs_*, custom keys) — not display labels.
Build a dictionary { hubspot_internal_name → primicornis_property_key }.
Coerce enumerations (e.g., deal stage) to Primicornis select options before bulk insert.

Relationship order

person, company / startup
Link people ↔ companies (roles like founder, board)
opportunity records with foreign relations to company + champion contact

Deduplication

Prefer HubSpot record id stored in a dedicated text property (hubspot_id).
Upsert strategy: IF EXISTS hubspot_id THEN PATCH ELSE POST.

CSV vs API migration

Mode	When
CSV	Initial bulk load, historic archives
API	Incremental sync, nightly delta pulls from HubSpot (private apps)

Gotchas

Multi-currency deals — normalize to workspace currency before import.
Marketing contacts — decide if they become person records or stay in a marketing object; mixing inflates CRM noise.
Automated sequences — disable HubSpot workflows during cut-over week to avoid ghost updates.

API-flavored sketch

bash

# Pseudocode: create startup with external ref
curl -X POST "$API/ws/$WS/object-definitions/$STARTUP_ID/records" \
  -H "Authorization: Bearer $PRIM_TOKEN" -H "Content-Type: application/json" \
  -d '{"properties":{"name":"Nova Labs","hubspot_company_id":"123456"}}'

Validate each batch with GET queries filtered by your external ID field before moving stages in pipeline imports.