The four cost drivers
| Driver | What moves it | Usually underestimated? |
|---|---|---|
| Item count | How many things actually get tagged — which should be fewer than the asset register suggests. | No. This is the number everyone starts with. |
| Tag type | Environment: laundry, metal, outdoor, chemical exposure, or benign indoor. The spread between a printed label and an industrial laundry or on-metal tag is large. | Sometimes — usually when a generic price was quoted before anyone looked at the assets. |
| Read points | Handhelds are cheap and flexible; fixed portals cost more and require power, mounting and sometimes civils; tunnel or conveyor reads more again. Our reader price guide puts numbers on each class. | Yes. And this is what determines what the system can actually see. |
| Tagging labour & register build | Physically attaching tags, recording what each one is, and reconciling that against whatever register exists today. | Yes — most of all. On a large population this frequently exceeds the hardware. |
| Integration | Whether the data lives in its own portal or flows into your PMS, ERP or stock system. | Yes, when the existing system's interfaces are discovered late. |
This is why we quote a cost band after a survey rather than a rate per tag. A per-tag price is easy to compare and tells you almost nothing: it hides the labour of putting the tags on, the design work that decides whether they are ever read, and the integration that decides whether anyone uses the result.
Choosing the tag: environment first, price second
UHF RFID has two physical enemies — metal reflects and liquid absorbs. Almost every tag-selection mistake traces back to ignoring one of them.
- Printed labels (paper or synthetic). Cheapest by a distance, ideal for cartons, files, stock on shelves and low-value indoor items. They fail on metal, in wet conditions and under abrasion.
- On-metal tags. Constructed with a spacer or ferrite layer so the antenna is not detuned by the surface. Necessary for tools, gas cylinders, racking, IT hardware, valves and instrument housings. They cost meaningfully more, and there is no way around that — a standard label on a steel tool is a decoration.
- Industrial laundry tags. Sewn or heat-sealed into textiles, built for repeated hot washing, detergent, bleach, extraction and pressing. Typical ratings start around 200 wash cycles, with premium thermoplastic types independently tested past 500 cycles at 95 °C while keeping read range. For linen and uniform pools, the cost question is not the tag price but the tag price divided by the item's remaining service life.
- Hard tags (screw-fix, cable-tie, rivet). For assets that live outdoors, get handled roughly, or must not have the tag removed casually. Ruggedness and tamper resistance are what you are paying for.
What tags actually sell for in South Africa
Published local retail pricing is thin, but it exists and it frames the conversation. South African electronics retailers list standard RFID tags and cards from around R10 to R275 each at single-unit retail (Netram's published range, September 2026), with basic 125 kHz access-type tags at the R10 end (MiRO lists one at R10.71 incl. VAT). Volume changes everything: plain UHF label inlays fall to a few rand each in project quantities, while specialised constructions — on-metal housings, industrial laundry tags, rivet-mount rugged tags — sit at the top of the range and above it, because you are buying mechanical engineering, not just a chip. If a quote prices every tag type the same, the quote has not looked at your assets.
Two practical rules follow. Test before you buy in volume — a short trial on your actual assets, in your actual environment, with your actual reader, settles more arguments than any datasheet. And expect a mixed population: a real project usually needs two or three tag types, because a store contains cartons, steel tools and a few outdoor items, and pretending otherwise produces a system that reads 80% of the time and gets abandoned.
When RFID is the wrong answer
An honest supplier should be willing to talk you out of it. Barcodes remain the better choice when items are handled one at a time anyway, when volumes are small enough that bulk reading buys nothing, or when the unit value is so low that any tag cost is disproportionate. Manual counting is fine when the population is small and stable. And where you genuinely need to know where a moving thing is right now — not where it was last seen — RFID is the wrong technology entirely, and something location-based is what you want.
The related question is what to tag. Tags cost money per item, so they belong on things that carry value, walk off, or raise a custody question. Tagging low-value consumables in the interest of completeness is the most reliable way to make a business case fail, and we would rather scope a smaller project that works.
Where the payback actually comes from
Three places, in the order we usually see them land:
- Loss that becomes visible. You cannot reduce a shrinkage rate you have never measured. The first accurate count is often the business case on its own.
- Time. A stocktake that took days takes minutes, and can then happen monthly instead of annually — which is what actually catches problems while they are small.
- Billing you can verify. Wherever you are charged per item counted by someone else — a laundry contract is the clearest example — an independent count changes the conversation permanently. That is the same reasoning we apply to metered water and to energy: whoever holds the count holds the argument.
How we quote it: a site survey and item census first, then a written cost band covering tags, read points, software, integration and the tagging work itself — because we deliver asset tracking as a turnkey project, the tagging labour is inside our scope rather than a surprise inside yours. If the survey says the numbers do not work, we will tell you that too; it is a cheaper conversation than a system nobody uses.