Retail RFID Inventory: Labels, Encoding & Pilot Checks

How a label becomes an inventory record

A reader observing an EPC is the start of an event. It does not by itself establish whether an item was received, sold, returned or moved. Define the intended workflow and assign an owner to each step:

1. Identify the item. Apply the approved label and encode its assigned identifier. Keep the item, printed barcode and EPC mapping together.
2. Collect observations. A compatible UHF reader reports EPCs with available reader, antenna and time information. Confirm the installation country, settings and physical read zone.
3. Interpret the event. Reader software or middleware groups repeated observations, applies zone and workflow rules, and sends exceptions for review. An item read repeatedly is still one item.
4. Update the business system. An agreed interface maps the item and event to inventory, POS, WMS or ERP records. Test acknowledgement, failed updates, retries and reconciliation with the system owner.

Start with one defined task: receiving against a known shipment, counting a sales-floor zone, finding a particular item, or verifying a return. A receiving portal and a staff-operated count can require different read zones, exception rules and acceptance tests. Choose the first deployment from the operating problem and evidence gathered in the pilot.

Phone NFC is a different buying route. A phone's built-in NFC reader does not read these UHF EPC Gen2 labels. A phone can be part of the workflow when paired with compatible external UHF hardware and software. For consumer taps to a website, evaluate NFC stickers and test the intended phone experience separately.

Check the retailer's category before choosing the inlay

A chip name is not a retailer approval. Auburn's ARC specifications consider the use case, product or packaging, environment and RFID infrastructure; approved inlays are assigned against those specifications. Obtain the requirement for the actual selling unit and category, then identify the exact proposed inlay on the appropriate list. See Auburn's explanation of ARC specifications.

For a Walmart project, start from the current Walmart category and inlay lookup at Auburn. Confirm the applicable requirement with the program owner instead of relying on a general list of covered categories or an old supplier presentation. A list entry for another category does not establish suitability for your product.

Then validate the finished, encoded sample and its placement. Where the retailer requires ALEC validation, Auburn's submission process checks the approved inlay, tagging location, encoding and other end-user requirements. Confirm the applicable submission instructions and keep the resulting status with the order record. Supplier availability and a successful desk read do not replace that process. See Auburn ALEC validation.

Keep a dated approval record: retailer and category, requirement version or retrieval date, inlay maker and model, label drawing, placement photograph, encoding example and validation reference. Recheck the relevant requirements when the item, packaging, label or inlay changes. For a US sourcing brief, use the US retail UHF label purchasing page.

Match the EPC scheme before choosing chip capacity

In an SGTIN, the GTIN identifies a trade-item class and the serial distinguishes individual units sharing that GTIN. Keep the tag's TID separate from the EPC used to identify the item. The GS1 EPC Tag Data Standard defines these encodings and memory roles.

SGTIN-96 and SGTIN-198 have different capacity and serial-format requirements. SGTIN-96 uses a numeric serial with defined limits; SGTIN-198 supports a longer alphanumeric representation and needs 198 bits for that EPC encoding. It cannot fit into a chip offering only 96 or 128 EPC bits. Obtain representative serial values and the required scheme before approving the part.

Use this table to check capacity and supply questions. It compares IC characteristics, not finished-label read performance or retailer approval.

Swipe the table to compare chip and order checks.

Confirm who owns identifier allocation, which data file is authoritative and how replacements avoid duplicate item identities. Read back an encoded sample and have the receiving system decode it. Compare the result with the intended item and visible barcode. Approve supported lock or password settings for the exact chip; do not apply permanent changes before this check.

Chip references: Impinj Monza R6 documentation, Impinj M700 specifications and NXP UCODE 8 specifications and manufacturing status.

Specify the finished label, reader and printer together

• Item and attachment: provide the label area, curvature, packaging layers, nearby metal, foil and contents. Test liquid-filled products at representative fill levels. Select a suitable UHF on-metal construction when needed; a standard sticker or NFC isolation layer is not an interchangeable solution.
• Finished construction: name the chip and inlay, then approve the label outline, face material, adhesive, liner and application position. Include handling and exposure requirements that need documented performance or testing.
• Reader setup: identify the country, reader model, antenna, settings, orientation, item density, movement and read window. A handheld result does not establish portal performance or a guaranteed range.
• Printer and encoder: confirm the exact printer model and installed UHF RFID option, supported inlay, firmware and software. Approve print stock, inlay position, label pitch, liner width, roll core, outer diameter and winding on a sample roll.
• Supply responsibility: distinguish blank labels, printed labels awaiting your encoding, and printed-and-encoded labels. Assign artwork, data allocation, write verification, reject handling and acceptance records to named teams.

The UHF sticker product page brings these label and roll requirements into one sample request. Reusing a familiar chip does not remove the need to approve a different antenna, adhesive, packaging or printer setup.

Run a pilot that can explain its failures

Use a physically checked test population and a saved expected-EPC list. Record the item mix and label coverage before testing; untagged items cannot appear in the RFID population. Include representative difficult items, intended stacks and orientations, and items just outside the target zone. Keep the original result when a retry is required.

1. Agree the protocol. Define the task, boundary, expected items, start and stop conditions, operator route and allowed retries. Freeze movements or log them so the physical reference remains meaningful.
2. Save observations and exceptions. Keep the reader configuration, raw observations and grouped unique EPCs. Separate expected items missed from unintended nearby items observed.
3. Verify identifiers physically. Investigate missing, unknown or mismatched records. Two labels bearing the same EPC are an encoding defect; repeated reads of one correctly identified item are normal observations that software must handle.
4. Follow the event to inventory. Check the right item and location, successful update, retry behavior, rejected events and manual correction history. Review a return or replacement if that operation is in scope.
5. Repeat after a controlled change. Change one relevant factor, such as placement or reader setting, retain the previous record, and compare the same measures. Agree the final acceptance criteria before production.

An illustrative pilot record: 200 items, not 1,250 reads

This is a hypothetical worked example, not an RFIDAK customer result or an acceptance promise. Assume a closed test zone contains 200 individually verified items, each with one different, correctly mapped EPC. No item moves during the test. The reader reports 1,250 observations, which software reduces to 199 distinct EPCs: 197 expected and two outside the expected list.

Swipe the table to read each measure and its meaning.

A read-rate percentage is not an inventory-accuracy percentage. Compare system stock with a separate physical audit at a defined time. For example, if your metric is exact SKU-location count agreement, its denominator is the number of SKU-location records audited. It is different from detecting expected EPCs. Log items without tags, duplicate-encoded items, reservations, damaged stock and movement exceptions rather than hiding them inside one favorable percentage.

Compare the full cost over the same operating scope

Request prices for the same label, data work, quantities per design and delivery terms. Then build the project budget separately. A lower label unit price does not establish a lower deployed cost.

Swipe the table to compare purchasing cost categories.

A useful planning formula is: project cost = setup and integration + equipment + recurring labels and encoding + operating and software costs + delivery costs, measured over one declared period. Compare it with the cost and measured outcomes of the existing process over that same period, item population and scope.

For labor, compare completed counting and reconciliation hours multiplied by your own loaded hourly cost. Include retry and exception time. A business case can also track verified stock corrections or fulfillment outcomes, but do not infer a sales or shrinkage benefit directly from read rate or count the same benefit twice.

What to send with a retail label RFQ

Send the retailer and selling country; current category or inlay specification; item and packaging photos; proposed attachment position; chip and inlay if specified; EPC scheme and sample data; printer and reader details; label and roll drawing; quantities per design; and destination. Share the pilot's failed cases as well as its best result so the supplier can assess the construction that needs to change.

RFIDAK's role in this request is label supply and the agreed printing, encoding and sample work. Confirm system design, installation, software integration and retailer validation responsibilities with the relevant teams. A label quotation should not imply that those services or approvals are included.

The standard tags-and-labels order baseline is 1,000 pieces. The quote confirms the exact configuration, quantities per design, availability, sample charges and production schedule; freight and arrival planning depend on the destination. Review the sample policy before agreeing the test shipment.