Short answer: the DESADV (Despatch Advice) is the EDIFACT message a supplier sends when goods actually leave — it is the EDIFACT counterpart of the X12 856 · Ship Notice / ASN. An ORDERS message says what the buyer wants; a DESADV says what physically shipped, in which packages, under which identifiers. If the DESADV disagrees with the truck, the receiving dock — not your system — finds out first.
I build EDI tooling for a living, and DESADV is the message where European supply chains quietly succeed or fail. Here is how it actually works on the wire.
What does a DESADV contain?
A DESADV is built from a small set of segments that repeat in a strict hierarchy:
- UNH/UNT — the message envelope: reference number, message type (DESADV), version (commonly D.96A or D.01B), and the segment count plus control reference that must match at the tail.
-
BGM — the document identity:
BGM+351+DA-88412+9'declares a despatch advice (code 351) with the sender's despatch number. Receivers key their goods-receipt posting off this number, so it must be unique per despatch — reusing it is the EDIFACT equivalent of sending two different 856 · Ship Notice / ASN files with the same shipment ID. - DTM — dates that matter: despatch date/time (DTM+11), estimated delivery (DTM+17). Format qualifiers in the 102/203/204 family decide whether the receiver reads your date as a date or as noise.
- NAD — the parties: NAD+BY (buyer), NAD+SU (supplier), NAD+DP (delivery party). The GLNs in these segments are how the receiver's system routes the advice to the right warehouse; a wrong delivery-party GLN sends your ASN to a building that is not expecting your truck.
- CPS — the consignment packaging structure: this is the hierarchy. CPS segments nest shipments into packages, and the PAC/PCI/GIN segments under each CPS describe the packaging and carry the SSCC (Serial Shipping Container Code) — the licence plate the dock scans.
Why is the hierarchy the hard part?
In a flat message, a line item is a line item. In a DESADV, every LIN (line item) hangs under a CPS node, which means the message asserts a physical claim: these items are in this box, this box is on this pallet. Retailers and automotive receivers validate that claim mechanically:
- Every SSCC in PCI/GIN must be unique within the message and check-digit valid.
- The quantities across the CPS tree must sum to the shipped quantities on each LIN — no more, no less.
- The packaging counts in PAC must match the number of CPS children that claim that level.
When any of these fail, the receiver does not send you a friendly email. You get a CONTRL (the EDIFACT syntax acknowledgment) if the file is malformed, or an APERAK (application error message) if it parsed but the content was wrong — or, worst of all, silence plus a chargeback when the goods arrive and the dock cannot match them to an advice.
That is the same failure shape as an X12 ASN rejected after the 997 · Functional Acknowledgment already said "accepted": syntax acceptance is not business acceptance. The DESADV equivalent of the 997 only proves the envelope opened.
What breaks most often in production?
Three things, in our experience:
Version drift. A partner certifies on D.96A, then their ERP upgrade quietly starts emitting D.01B segment order. Your map "works" in testing because the test file was last year's. Pin the version per partner, and treat a UNH version change as a breaking API change — because it is one.
SSCC reuse. Warehouse systems that generate SSCCs from a local counter collide after a restore or a site migration. The receiver sees a duplicate licence plate and rejects the whole consignment, not just the pallet. Generate SSCCs from a single durable sequence, and never recycle one inside the retention window your partners scan against.
Quantity truth. The DESADV is often generated from the order or the pick plan instead of the actual pick. Short-pick one case and the advice overstates the shipment; the receiver books goods receipt against the DESADV, then the invoice (INVOIC, the counterpart of the 810 · Invoice) matches the advice instead of the shelf — and now finance owns a logistics bug. Generate the DESADV from the confirmed pick, after the pallet is wrapped, not before.
How should you test a DESADV flow?
Test the tree, not the segments. A file where every segment is individually valid but the CPS counts disagree with the LIN quantities is the most common real-world failure, and segment-level validators wave it through. Your test set needs at least: a single-package shipment, a multi-pallet mixed shipment, a short-pick, and a duplicate-SSCC negative case that must fail loudly on your side before it ever leaves.
If you are bridging worlds — say your ERP thinks in X12 because your US partners do — remember the DESADV and the 856 · Ship Notice / ASN carry the same business truth in different grammars. Map the hierarchy first and the segments second; teams that map segment-by-segment produce flat ASNs and nested DESADVs that quietly drop the pallet level.
FAQ
Is DESADV the same as an ASN? Yes in purpose: DESADV is the EDIFACT despatch advice; the ASN is the X12 856 · Ship Notice / ASN. Both announce a shipment with package-level detail before arrival.
What acknowledges a DESADV? CONTRL for syntax, APERAK for application-level errors. Neither confirms the goods matched — only goods receipt at the dock does that.
Which segment carries the SSCC? The GIN segment under PCI at each packaging level of the CPS hierarchy.
We handle DESADV and 856 traffic side by side at SignalEDI, including the hierarchy validation that catches count mismatches before a partner does. I'm Chris, founder of SignalEDI — if your despatch advices keep passing syntax and failing docks, the tree is where I'd look first.
Top comments (0)