# FSMA 204 KDEs From Suppliers Who Do Not Send EDI > Half your supplier base will still be emailing PDFs in 2028. How to capture FSMA 204 Key Data Elements from invoices, packing lists, and faxes without hand-keying every one. Suppliers who do not send EDI still send the data, just on paperwork: PDF invoices, emailed packing lists, faxed bills of lading. FSMA 204 requires records in an electronically sortable format within 24 hours, so that paperwork has to become structured data. The realistic options are supplier pressure, portal entry, or automated document extraction. Most FSMA 204 readiness advice assumes a supply chain that already trades electronically. Buy the traceability platform, connect it to the ERP, map the [856 ASN](/blog/fsma-204-edi-856-shipping-cte), done. Then you look at your actual supplier list and find that a third of your Food Traceability List volume comes from growers, co-packers, and specialty producers who have never sent an EDI document and are not going to start because of an FDA rule aimed at you. This is the part of FSMA 204 readiness that does not solve itself, and it is worth being specific about it. ## Why the gap concentrates exactly where you do not want it There is an unhappy correlation here. The suppliers least likely to be EDI-enabled are small growers, regional produce houses, cheesemakers, egg producers, and seafood suppliers. Those are also, disproportionately, the people handling the categories on the Food Traceability List. Your national branded suppliers are already sending clean ASNs. Your traceability exposure is concentrated in the half of the supplier base that emails you a PDF. ## What the Receiving CTE actually needs If you are a distributor, wholesaler, or retailer, Receiving is your Critical Tracking Event. For each traceability lot you take in, you record: 1. The Traceability Lot Code 2. Quantity and unit of measure 3. Product description 4. The immediate previous source, meaning who shipped it to you 5. The location where you received it 6. The date you received it 7. A reference document type and number Six of those seven are already on an ordinary invoice or packing list. Supplier name, product, quantity, date, your receiving DC, the PO or invoice number. The one that frequently is not present is the Traceability Lot Code, and that one is the point of the entire rule. So there are really two problems, and it helps to keep them apart: - **A data problem**: is the lot code on the document at all? - **A format problem**: even when everything is on the document, a scanned PDF is not an electronically sortable record. ## Option 1: push the supplier This is the correct first move and it costs nothing but time. Ask each in-scope supplier when they will begin putting a Traceability Lot Code on shipping paperwork, and get a date. GS1 US guidance is clear that operational documents like ASNs, bills of lading, invoices, and case labels should carry the TLC for in-scope items. Be realistic about the hit rate. A supplier who ships to three distributors will do it. A supplier who ships to sixty and is already fielding sixty different data requests will deprioritize yours. Track the responses, because the ones who say no are your remaining scope. Note also that pushing a supplier to print the lot code solves the data problem and does nothing for the format problem. You still have a PDF. ## Option 2: manual entry at receiving Someone at the dock or in the back office keys the KDEs into a portal or spreadsheet as product arrives. This genuinely works at low volume, and for a distributor with a handful of FTL SKUs it may be the right answer. Do not build a system for a problem you do not have. It stops working on volume and on accuracy. Lot codes are long alphanumeric strings with no check digit and no validation, typed by someone under time pressure at a receiving dock. A transposed character produces a record that looks complete, passes every review, and fails at exactly the moment you need it: the trace-back. The failure mode is silent, which is what makes it dangerous. ## Option 3: automate the document The third option is to treat the inbound document as the data source it already is, and convert it automatically. The supplier keeps emailing a PDF. The PDF gets read into structured line-level records. Your traceability system receives data instead of an attachment. This is what OrderSync does, and it is worth being precise about the boundary, because the space is full of vendors implying they solve more than they do. **What we do.** OrderSync ingests inbound orders and invoices however they arrive: [PDF](/pdf-order-processing), [email body](/email-order-processing), [fax](/fax-order-processing), CSV, or EDI. It extracts them into structured line-level records and keeps a full audit trail of the original document, what was extracted, what a human corrected, and when it synced. Which fields get captured is configurable per trading partner, so the fields a given supplier prints can be mapped to the fields your downstream system expects. **What we are not.** We are not a traceability system. We do not hold your lot history, we do not model Critical Tracking Events, and we will not be the thing that answers an FDA trace-back request. That ledger belongs in your food safety platform or your ERP, alongside vendors who specialize in it. We are the layer that gets clean data into it. If you are evaluating us against a traceability platform, you are comparing the wrong two things. The honest comparison is: what is your plan for the documents that arrive as paperwork, and is that plan "someone retypes them"? ## A way to size your own gap Before buying anything, get one number. Take last quarter. For the SKUs on the Food Traceability List, what percentage of receipts arrived with a usable lot code in structured form? - **Above 80%**: your gap is supplier follow-up on a short list. Handle it with procurement, not software. - **40% to 80%**: you have a real intake problem and two years to fix it. Automating document extraction is likely cheaper than the headcount alternative. - **Below 40%**: the document side is your FSMA 204 project, whatever else you buy. That single percentage is more useful than most readiness assessments, and you can get it from your receiving records in an afternoon. ## What to do next Start with the corrected timeline in our [FSMA 204 compliance guide](/fsma-204). The deadline is July 20, 2028, not January 2026, and a surprising amount of published guidance still has this wrong. If your suppliers do send EDI, the [Shipping CTE to 856 mapping](/blog/fsma-204-edi-856-shipping-cte) shows where the lot code belongs, and you can check what a real file contains with our [free EDI Inspector](/edi-inspector). If they do not, look at how [multi-format order processing](/multi-format-orders) handles the documents you are getting today. ## Sources - [FDA: FSMA Final Rule on Requirements for Additional Traceability Records for Certain Foods](https://www.fda.gov/food/food-safety-modernization-act-fsma/fsma-final-rule-requirements-additional-traceability-records-certain-foods) - [GS1 US: standards and guidance for food traceability](https://www.gs1us.org/industries-and-insights/standards) ## FAQ ### Does FSMA 204 require EDI? No. The rule is technology neutral. It requires that records exist, contain the required Key Data Elements, and can be produced in an electronically sortable format within 24 hours of an FDA request. A spreadsheet that genuinely meets those conditions is compliant. EDI is popular because it satisfies them as a byproduct of trading. ### What if a supplier refuses to provide a Traceability Lot Code? Document the ask and the response, then decide commercially. For FTL items you are the one carrying the receiving obligation, so a supplier who will not provide a TLC is transferring risk to you. Some distributors have made TLC provision a condition of purchase orders for in-scope categories. ### Can OCR alone solve this? Partly. Reading characters off a page is the easy half. The hard half is knowing that a particular string is a lot code rather than a PO number, that a quantity is cases rather than eaches, and that this supplier writes dates day-first. That mapping is per-partner work, which is why template-free extraction with a correction loop tends to beat generic OCR on real supplier paperwork.