The challenges of food traceability in 2026

A food product batch without complete traceability is not just a regulatory risk — it is a time bomb in your commercial margin. With wholesale margins structurally narrow in Portugal, a badly managed market recall consumes months of profit in a single event. This guide gives you a concrete process and an 18-point checklist to audit food traceability in your distribution operation — usable in tomorrow's meeting.

The thesis nobody says out loud

Most Portuguese food distribution companies believe they have traceability. They have records. It is not the same thing.

Paper records, Excel spreadsheets shared by email and generalist ERP modules that were never configured for food batches create an illusion of compliance. When an ASAE audit arrives, or a recall request from a large customer — a retailer with 80 stores demanding a response within 4 hours — the team discovers that the record exists but cannot be consulted in real time, does not link supplier to batch to end customer, and does not cover returns.

The 2026 problem is not the absence of data. It is the fragmentation of the data that already exists.

What you need before you start

Before touching any system or process, gather seven elements. Without them, any internal audit produces a beautiful, useless report.

First, the complete mapping of inbound flows: supplier, carrier, receiving warehouse — including suppliers who deliver directly to stores without passing through the central warehouse, who are systematically forgotten. Second, the list of product categories subject to mandatory traceability under Regulation (EC) No 178/2002 — do not assume you know it by heart; check with the quality manager. Third, the formal identification of who has the authority to stop a dispatch: in a distribution operation with rotating shifts, that authority cannot depend on a single person being available on the phone.

Fourth, access to the current management system — ERP, WMS or spreadsheet — with read permissions for batch history. Fifth, the maximum response time required by your customers in the event of a recall: the standard among large retailers is 4 hours for identification, 24 hours for physical recall. Sixth, the labelling and ATCUD requirements applicable to outbound documentation (DL 28/2019). Seventh, the inventory of cold-chain break points — and who records them today, how often, and in which system.

Step 1 — Define what a "batch" is in your operation

It sounds basic. It is not. In INFOS projects we see food distributors with three batch definitions coexisting simultaneously: the supplier's (printed on the packaging), the internal one (generated in the warehouse on receipt) and the customer's (who requests their own code in the EDI). When there is a non-conformity, the three systems do not talk to each other — and the team spends the first two hours of the crisis reconciling identifiers, not recalling product.

Decide on a single hierarchy before touching any system. Adopt the supplier's batch as the primary identifier wherever possible. Create an internal batch only when the product is split or repackaged — and in that case, always keep the source batch associated. Build an equivalence table of supplier-batch ↔ internal-batch ↔ customer-batch in your SCM or ERP, and document the rule in writing with the signature of the quality manager and the warehouse manager.

• ☐ There is a written and approved definition of "batch" in the company.
• ☐ The definition is configured in the management system.
• ☐ All warehouse operators know the rule.

Step 2 — Audit the data capture points

Walk the goods flow physically: receiving → storage → picking → dispatch → returns. At each point, answer three questions: is the batch recorded here, or merely assumed? Is the record made in real time or reconstructed at the end of the shift? If this point fails, is the next batch in the chain left without a traceable origin?

The most critical point — and the most frequently ignored — is the return. A product returned by a customer enters the warehouse without an identified batch in more than half the operations we audit internally. That product can be reintegrated into stock and dispatched again to another customer. Traceability breaks there, not at receiving. The warehouse operator is not to blame: if the system allows a return entry to be closed without the batch field filled in, they will close it — because there are another 40 pallets waiting.

Food traceability almost always fails in the same place: at returns and at splitting. These are the two moments when the original batch disappears from the record and nobody notices — until it is too late.

• ☐ Receiving: batch recorded by barcode or QR scan, not by manual entry.
• ☐ Picking: the picking document includes the batch to be collected, not just the reference.
• ☐ Dispatch: the delivery note or invoice includes the dispatched batch.
• ☐ Returns: there is a mandatory batch identification process before reintegration into stock — and the system blocks reintegration if the field is empty.

Step 3 — Test upstream and downstream traceability

Run two simulation exercises now, without warning the team. Do not wait for the real audit to discover them.

Upstream test (trace-back): pick a batch dispatched last week. Reach the supplier, manufacturing date and original batch number in under 15 minutes, without calling anyone. If you need to call the warehouse to ask where the receiving file is, the test has failed.

Downstream test (trace-forward): pick a batch received 30 days ago. Identify all the customers who received product from that batch, the volumes and the delivery dates — in under 15 minutes. If the result depends on cross-referencing three Excel files exported from different systems, the test has also failed.

If either exercise fails or exceeds the time, you have an operational gap, not merely a software gap. The distinction matters because the solution is different: an operational gap requires process redesign first, then technology. Reversing the order is the most expensive mistake we see in WMS implementations at food distributors.

• ☐ Trace-back completed in under 15 minutes without manual intervention.
• ☐ Trace-forward completed in under 15 minutes without manual intervention.
• ☐ Result documented and shared with management.

Step 4 — Check regulatory coverage

Regulation (EC) No 178/2002 requires "one step back, one step forward" traceability for all operators in the food chain. It does not require full chain traceability — but it does require each link to know its immediate supplier and its immediate customer for each batch. It is a minimalist requirement that most companies cite correctly and comply with poorly in practice.

Two vectors add pressure in 2026. The first is the growing demand for data transparency from large retailers, who request access to batch data via EDI or B2B portal — not as a regulatory requirement, but as a commercial condition. Refusing means losing the customer. The second is the European Commission's Sustainable Food Chain Strategy (Farm to Fork), which anticipates environmental traceability requirements — geographical origin, carbon footprint per batch — not yet mandatory but already demanded by large-scale buyers. Those who do not have the data architecture ready will pay double to build it in a hurry.

• ☐ Compliance with Regulation (EC) No 178/2002 documented and verified by legal or a quality consultant.
• ☐ Ability to respond to EDI requests for batch data from retail customers.
• ☐ Plan (even if 18 months out) to capture geographical origin data per batch.

Audit checklist — 18 points

Common mistakes and how to avoid them

Mistake 1 — Confusing reference with batch. The ERP has the product reference but not the batch. The warehouse operator assumes they are equivalent — and for years nobody corrected it because there was never a recall. The remediation is surgical: configure the system to make the batch field mandatory in all inbound and outbound transactions, without exceptions and without the possibility of the supervisor overriding it. If the field can be skipped "in emergency situations", it will be skipped regularly.

Mistake 2 — Traceability only forward. The company knows who it dispatched to, but cannot identify which supplier and batch the product came from. This mistake is more common in distributors that grew by acquiring regional warehouses, where each unit inherited its own receiving system. The remediation involves requiring the supplier to provide the batch number on the invoice or delivery note — and configuring receiving not to close without that field filled in. If the supplier does not include the batch on the document, receiving does not proceed.

Mistake 3 — Batch data in a system separate from the ERP. Quality uses an Excel spreadsheet, the warehouse uses the WMS, invoicing uses the ERP. None of the three talk to each other. In an audit, the team can prove traceability — but it takes three hours and involves four people. That is not operational traceability; it is documentary archaeology. The remediation requires integration of the three flows into a single data system — such as KORA Inventory Suite, which links warehouse capture directly to the ERP, eliminating manual reconciliation between systems.

Mistake 4 — Testing traceability only when there is a problem. The market recall procedure exists on paper but has never been simulated. When the real situation arrives, the team discovers that the quality manager changed eight months ago, that access to the batch history is locked by a password only the former IT person knew, and that the contact list of affected customers is in an Excel file on the computer of someone who is on holiday. Simulate a complete recall once a year — with a stopwatch.

A note on checklist point 17, which seems bureaucratic but is not: in distributors with 30 to 80 employees, access to the batch history frequently depends on a single person — the IT manager or the most senior warehouse manager. When that person is absent during a recall crisis, the operation stops. Distribute access permissions and document the consultation procedure in such a way that any member of the management team can carry it out autonomously.

The difference between a food distribution operation with real traceability and one with apparent traceability is not in the software installed. It is in knowing, at three in the morning on a Friday, which customers received batch 2024-11-07-A — and having the complete list in under a quarter of an hour, without waking anyone.

Frequently asked questions

What is considered a "batch" in food traceability in Portugal?

A batch is an identifiable set of units of a food product produced, processed or packaged under essentially identical conditions. In distribution, it can be the supplier's batch (printed on the packaging), an internal batch (generated at receiving) or the batch requested by the customer. It is critical to define a single, documented hierarchy to avoid confusion during market recall crises.

What is the maximum time to respond to a product recall request?

Large Portuguese retailers typically require 4 hours to identify the affected product and 24 hours for physical recall. This deadline is non-negotiable and requires a traceability system that can be consulted in real time, without the need to contact third parties or consult scattered files.

Why does traceability fail at returns?

The return is the most ignored critical point. Returned products frequently enter the warehouse without an identified batch and can be reintegrated into stock without a record of origin. If the system allows a return entry to be closed with the batch field empty, the operator will do so — because of work pressure. Traceability breaks there, not at receiving.

Which regulation requires food traceability in Portugal?

Regulation (EC) No 178/2002 establishes the mandatory traceability requirements for food products. In Portugal, Decree-Law 28/2019 complements these requirements with specifications on labelling and ATCUD (Transport Authorisation for Ungulate Meats). Each product category may have specific requirements.

How do you test whether traceability is actually functional?

Carry out two tests without warning the team: trace-back (pick a batch dispatched a week ago and reach the supplier in under 15 minutes) and trace-forward (pick a batch received 30 days ago and identify all the customers who received it in under 15 minutes). If you need to make phone calls or cross-reference Excel files, the system has failed.

What is the financial impact of a badly managed market recall?

With wholesale margins in Portugal structurally narrow, a badly managed market recall can consume months of profit in a single event. In addition, it incurs regulatory fines, loss of customer confidence and possible suspension of supply. Efficient traceability drastically reduces the time and cost of response.

What does "data fragmentation" mean in food traceability?

Fragmentation occurs when the data exists but is scattered: paper records, Excel spreadsheets shared by email, ERP modules not configured for batches. This creates an illusion of compliance. The real problem is not the absence of data, but the impossibility of consulting it in real time, linking supplier to batch to end customer in a traceable way.

Sources

• Regulation (EC) No 178/2002 of the European Parliament and of the Council of 28 January 2002, laying down the principles and requirements of food law, including mandatory traceability
• Decree-Law No 28/2019 of 28 February, transposing Directive (EU) 2017/625 on official controls and the enforcement of legislation on food and feed, including labelling and ATCUD requirements
• Statistics Portugal (INE) — Wholesale Trade Statistics and Commercial Margins in Portugal (2024)
• Food and Economic Safety Authority (ASAE) — Audit and Compliance Guides on Food Traceability
• ISO 22005:2007 — Traceability in the feed and food chain — General principles and basic requirements for system design and implementation