[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"glossary-field-mapping::en":3,"gloss-cluster-field-mapping::en":23,"gloss-next-field-mapping::en":9},{"slug":4,"category":5,"name":6,"definition":7,"meta_desc":8,"faq":9,"schema_markup":9,"related":10},"field-mapping","integration","Field Mapping","Field mapping is the configuration that declares which field in one system corresponds to which field in another — this CRM's Company Name is that billing tool's Account, this status value means that stage. It is the unglamorous middle of every integration and reliably the part that takes the longest, because two products that both model customers almost never model them the same way. The mismatches are structural rather than cosmetic. One system stores a single full name where the other keeps first and last separately. One has a fixed picklist of five deal stages, the other free text. One treats a company as the primary record with contacts hanging off it, the other treats the person as primary. Formats differ on dates, currencies, phone numbers, and country codes; required fields on one side may be optional on the other, so records that are perfectly valid in the source are rejected on arrival. Custom fields, which are where most of a company's actual business logic ends up living, are rarely covered by a prebuilt connector at all. Transformations are what bridge the gap: splitting or concatenating values, mapping one vocabulary of statuses onto another, defaulting a value the source does not carry, and skipping records that cannot be mapped cleanly rather than writing them wrong. Whether a tool supports transformations, and whether they can be expressed without writing code, is a much better question to ask in an evaluation than how many integrations the marketplace lists. Two practical habits pay off: keep the mapping documented somewhere outside the vendor's UI, because it encodes decisions nobody will remember in a year, and validate a mapping against a sample before running a full backfill, since a wrong mapping applied to history is far more expensive to unwind than one caught on twenty records.","Field mapping declares which field in one system matches which in another. It is the slow middle of every integration, and custom fields are rarely prebuilt.",null,[11,14,17,20],{"slug":12,"name":13},"backfill","Backfill",{"slug":15,"name":16},"integration-marketplace","Integration Marketplace",{"slug":18,"name":19},"sync-conflict-resolution","Sync Conflict Resolution",{"slug":21,"name":22},"two-way-sync","Two-Way Sync",[24,28,31,34,37,41,42,45,48,51,54,57],{"slug":25,"category":5,"name":26,"updated_at":27},"backend-for-frontend","Backend for Frontend (BFF)","2026-08-24T02:46:38+00:00",{"slug":29,"category":5,"name":30,"updated_at":27},"concurrency-limit","Concurrency Limit",{"slug":32,"category":5,"name":33,"updated_at":27},"event-ordering","Event Ordering",{"slug":35,"category":5,"name":36,"updated_at":27},"function-schema","Function Schema",{"slug":38,"category":5,"name":39,"updated_at":40},"grpc","gRPC","2026-08-24T02:46:37+00:00",{"slug":15,"category":5,"name":16,"updated_at":27},{"slug":43,"category":5,"name":44,"updated_at":27},"ip-allowlist","IP Allowlist",{"slug":46,"category":5,"name":47,"updated_at":27},"json-web-token","JSON Web Token (JWT)",{"slug":49,"category":5,"name":50,"updated_at":27},"mcp-server","MCP Server",{"slug":52,"category":5,"name":53,"updated_at":27},"mutual-tls","Mutual TLS (mTLS)",{"slug":55,"category":5,"name":56,"updated_at":27},"oauth-scopes","OAuth Scopes",{"slug":58,"category":5,"name":59,"updated_at":27},"openapi-specification","OpenAPI Specification"]