Recognising the failure pattern
Standard NFC tags contain a coiled antenna that communicates with a reader by generating a small alternating magnetic field. When that antenna sits directly against a metal surface, the metal does not simply block the signal in the way a wall blocks Wi-Fi. Instead, it becomes an active participant in the electromagnetic exchange. The metal surface generates opposing eddy currents that cancel out much of the tag's field, effectively detuning the antenna and collapsing the read range.

In practice, a tag that reliably reads at three to five centimetres in free air may drop to under one centimetre, or fail entirely, when pressed flat against steel, aluminium, or even certain metallised finishes. The effect varies with the metal type, thickness, and the tag's own antenna design, but the outcome is predictable: a standard tag on an unprepared metal surface will underperform or stop working.
The correct solution is a purpose-made on-metal tag. These incorporate a ferrite layer between the antenna and the adhesive backing. The ferrite acts as an insulating barrier for the magnetic field, preventing the eddy currents from forming. On-metal tags are slightly thicker and more expensive than standard tags, but they are the only reliable fix for direct metal contact. Spacing a standard tag away from the surface with foam tape or a plastic mount can restore some range, but the required gap depends on the tag, the metal, and the reader, making it an unreliable approach for a deployed installation.
A structured path from symptom to cause
This problem turns up most often in environments where metal is part of the fixture rather than an obvious hazard. A few typical scenarios:
- Museum display cases. Many cases have steel or aluminium frames, and curators sometimes stick a tag to the frame for convenience. If the case itself is metal, even a tag placed on the glass near the metal edge can suffer reduced range.
- Retail shelving. Gondola ends, wire baskets, and refrigerated cabinets are routinely metal. Tags applied to the shelf edge for product information or stock lookup frequently fail unless on-metal variants are specified.
- Event and exhibition signage. Shell schemes, pop-up stands, and rigging frames are often aluminium. An integrator who tests tags on a wooden desk before deploying them on a metal frame will discover the problem on the show floor.
- Industrial and warehouse assets. Racking, tool cribs, and machinery housings are obvious metal surfaces, but the mistake still occurs when procurement orders standard tags in bulk without checking where they will be mounted.
- Hidden metal behind surfaces. A plasterboard wall with metal studs, a glass panel with a steel backing, or a laminated surface with a metallic core can all produce the same detuning effect even though the surface itself does not look or feel like metal.
In each case, the operational pattern is similar: tags are tested in an office or on a non-metal sample, they read well, and the problem only appears once the full batch is installed on site. By that point, the labour cost of removal and replacement has already been incurred.
When a Spacer Might Be Acceptable
If on-metal tags are not available and a small number of standard tags must be used temporarily, a non-conductive spacer of roughly five millimetres or more can partially restore readability. Air, foam, acrylic, and thick card all work better than thin tape. However, the restored range will still be shorter than a dedicated on-metal tag, and the spacer adds a visible, tamper-prone element to the installation. Treat spacers as a short-term workaround, not a specification.
How to know the issue is genuinely resolved
The most frequent error is assuming that a tag which reads when a phone is pressed hard against it is working acceptably. In a real visitor or customer scenario, people will not press firmly or hold their device at a precise angle. If the tag only reads with deliberate, close contact, the installation will generate complaints and low engagement, even if it technically functions in a bench test.
Other recurring mistakes include:
- Ordering one tag type for the whole site. Mixed environments — some metal fixtures, some wood, some glass — require mixed tag types. A single SKU approach saves a few pence per tag but creates a class of silent failures on metal surfaces.
- Confusing water and metal interference. Water and metal both degrade NFC performance, but for different reasons and with different mitigation. A tag that fails on a metal drinks cooler is not failing because of condensation.
- Ignoring the adhesive backing. Some on-metal tags ship with adhesive that is not rated for the temperature range or surface finish of the target environment. A tag that works electrically but falls off after a week has not solved the problem.
- Not testing with the actual reader devices. Different phone models have different NFC antenna positions and sensitivities. A tag that reads on one handset at two centimetres may fail on another at the same distance, particularly when the metal surface is already reducing the margin.
Pre-Deployment Checks
Before committing to a full installation, confirm the following on site:
- Identify every surface the tag will contact. Check for metal directly, and probe for hidden metal behind cladding or glass using a magnet or a stud finder.
- Test a sample of the exact tag model, with its intended adhesive and any planned spacer, on the actual surface. Use at least two different phone models.
- Measure the reliable read distance — the distance at which the tag reads consistently without careful aiming — rather than the maximum distance under ideal hand placement.
- If the reliable read distance is below one centimetre on metal, switch to an on-metal tag before proceeding.
- Document which tag type goes where in your asset register, so that future replacements use the correct variant without guesswork.
For guidance on building that asset record, see the companion material on labelling and inventorying beacons and tags. For broader questions about NFC behaviour, the FAQ in this section covers common points of confusion.

