Asset Management 7 min read

The stocktake that takes minutes changes more than the stocktake.

When a count stops costing a weekend, you count monthly. When you count monthly, problems surface while they are still small. That — not the scanning trick — is what RFID changes in a stockroom.

By Frank Guo · Technology & Product Leadership, addanode

TL;DR — An RFID stocktake works because UHF tags answer in bulk without line of sight: one person with a handheld sweeps a stockroom in minutes, reading through boxes and packed shelves. The prize is not the faster annual count — it is count frequency: weekly or monthly cycle counts that catch receiving errors, mis-picks and shrinkage while the trail is warm. South Africa's largest fashion group has run an item-level RFID programme since 2019, and global item-level tag pricing has fallen to cents in volume — but the economics still turn on who applies the tags: source-tagged stock (tagged by the supplier) makes the case easily; tagging everything yourself in-store is where business cases go to die. Start with the categories that shrink, not the whole range.

Why counting is the constraint

Ask any South African stockroom what it knows about its inventory and the honest answer is "what the system says, minus whatever has gone wrong since we last counted". Barcode counts are accurate but slow — every item handled, every label found and faced — so full counts happen annually, cycle counts happen when there is time, and the gap between the system and the shelf grows quietly in between. Every downstream decision — replenishment, online availability, the shrinkage number the auditors see — inherits that gap.

RFID attacks the constraint directly. Because passive UHF tags respond in bulk, through cardboard and fabric, at metres of range, the count decouples from item handling entirely. The physical work of knowing what you have drops by an order of magnitude — and once counting is cheap, you do it often, which is the actual value.

What changes operationally

Cycle counts become routine. The count that took a team a weekend takes one person part of a morning. Weekly counts catch the receiving error this week, not at year-end when the trail is cold.
Receiving verifies itself. A tagged carton is counted as it arrives — against the ASN, in seconds, without opening it. Supplier short-shipments stop being discovered three months later.
Shrinkage gets a timeline. Frequent counts turn "we lost 2% this year" into "we lost these items, in this window, from this area" — which is the difference between a write-off and an intervention.
The system number becomes usable. Online availability, click-and-collect promises and replenishment triggers are only as good as stock accuracy; frequent RFID counts are what keep that number honest between annual audits.

The operational change above, column by column. The prize is the frequency, not the speed of any single count — and as the section below says, the economics turn on who applies the tags rather than on what a tag costs.

The economics, honestly

Item-level tagging only pays under conditions, and a straight supplier should name them. It works when tags arrive on the stock — source-tagged by the supplier or applied once at your DC — rather than being applied item by item in-store; when the category has real shrinkage or real availability cost, so a percentage-point of accuracy is worth money; and when the item value comfortably carries a tag worth a few rand. It does not pay for tagging low-value lines in the name of completeness — the same discipline we apply in asset tagging. The scale precedent is public: TFG, one of South Africa's largest fashion groups, has run an item-level RFID rollout since 2019, reporting cost savings and significant cuts in stock counting time. The direction of travel in tag pricing is equally public — item-level UHF labels cost cents in project volume. What is not public anywhere in South Africa is system pricing, which is a gap we address in our reader price guide.

Start smaller than the vendor deck suggests

The rollouts that stick start with one category and one question: the stockroom where shrinkage hurts, the line where availability drives sales, the bonded store where the audit costs real money. One handheld, tagged stock for that category, and a month of cycle counts produces the accuracy delta on your own floor — which is the only business case worth signing. Portal readers on receiving doors, dispatch lanes and yard gates extend the same tag population when the counts prove out. We deliver it turnkey — tags, readers, integration into your stock system, and the pilot design that tells you the truth before you scale.

FAQ

RFID stocktakes — common questions

How much faster is an RFID stocktake than barcode counting?

An order of magnitude, because the work changes shape: instead of handling every item and finding every label, one person sweeps a handheld through the area and every tag in range answers — through cartons, through packed rails. A stockroom count that took a team a day becomes a morning's walk. The bigger gain is that counts this cheap happen weekly instead of annually.

Does RFID reading work through boxes and packaging?

Yes — UHF RFID needs no line of sight, so tags read through cardboard, plastic and fabric. The physical limits are metal and liquid, which reflect and absorb the signal: dense stacks of foiled or liquid product need read-strategy design rather than a bigger reader. That is survey work, and it is why we pilot on your actual stock before scaling.

What do item-level RFID labels cost?

In project volumes, plain UHF label inlays cost cents to a few rand each — pricing has fallen steadily for a decade and is publicly tracked internationally. The tag is rarely the deciding cost. Who applies it is: source-tagged stock makes the arithmetic easy, while in-store tagging of every item adds labour that sinks weak business cases. Start with categories where suppliers can tag at source or where one tagging pass at the DC covers it.

Are South African retailers actually using RFID?

Yes — the public reference point is TFG, one of the country's largest fashion groups, which began an ambitious item-level RFID rollout in 2019 and reported cost savings and significantly faster stock counting. Adoption elsewhere in SA retail and warehousing is growing quietly; most operators do not publish their programmes, which makes the visible precedent more useful than the adoption statistics.

Do I need portal readers at every door?

Not to start. The highest-value first step is usually a handheld and tagged stock in one category — that alone delivers the cycle-count gain. Fixed portals on receiving, dispatch or gates come second, when you want movements captured automatically rather than counted periodically. Scaling read points after the pilot proves accuracy is cheaper than discovering mid-rollout that the strategy needs changing.

Can RFID counts feed our existing stock system?

Yes, and they should — a count that lives in a separate portal changes nothing. We integrate count and movement events into the stock or ERP system you already run, and hand over the reader APIs and documentation so the data path belongs to you. The count is only worth doing if it lands where replenishment and write-off decisions are made.

Primary sources

Just Style — "South Africa's TFG deploys RFID to drive stock accuracy" (September 2019; the 24-month rollout and reported results)
GS1 — EPC UHF Gen2 air interface protocol (the standard item-level retail tagging runs on)

Pick the category that hurts. Count it weekly.

Tell us where the stock number and the shelf disagree. We will design the pilot that measures the gap — and the system that closes it.

Scope a Project RFID asset tracking →