When is NFC the right choice for inventory?
When you go to the item and need certainty about which one you are looking at. For counting many items at once from a distance it is the wrong technology entirely, and a printed code is cheaper wherever the label stays clean and readable.
Where it is right
When someone is standing at the item anyway, for maintenance, inspection, picking or repair. The tap costs nothing on top of a visit that was happening.
When the label lives in conditions that destroy print: oil, abrasion, paint, weather, cleaning. A chip under a durable surface survives what ink does not.
When there is no line of sight, because a tag reads through a cover, a bag or a layer of dirt that would defeat a camera.
And when the record must be certain. A tap identifies one item unambiguously, where a person reading a number can transpose two digits.
Where it is wrong
Counting a room. NFC reads one tag at contact, so counting a hundred items means a hundred deliberate taps and a hundred walks.
Reading through a portal or a doorway. That is what a different radio technology at a different frequency exists for, and no tag size or reader changes it.
Tracking movement without a person. A tag reports nothing on its own, so every record requires somebody to arrive and tap.
Anything sold on the promise of automatic visibility. NFC gives you checkpoints where people already go, and nothing between them.
Against a printed code
A code costs nothing to produce, needs no chip, and works with any camera. That is a strong position and it wins wherever the label stays clean and visible.
NFC wins where the label does not stay clean, where there is no line of sight, where gloves and awkward angles make aiming a camera hard, and where the identifier must be rewritable.
It also wins on certainty in poor light, since a tap works in the dark and a camera does not.
Plenty of deployments use both on the same label, which costs a little and removes the argument entirely.
Deciding honestly
Ask whether anyone is going to the item. If nobody is, NFC records nothing and the project is about a different technology.
Ask what destroys labels in your environment, because that is the strongest argument for a chip and it is measurable rather than theoretical.
Ask what the count is actually for. Auditing a fixed asset register once a year is a different problem from knowing stock levels hourly.
And price a printed code first. If it would work, it is the cheaper answer, and knowing that before you start is worth more than a pilot that proves it late.
Frequently asked questions
Can NFC count a whole warehouse quickly?
No. It reads one tag at contact, so counting means walking to every item. Bulk counting at a distance is a different technology.
Is NFC better than a barcode?
Where labels get dirty, where there is no line of sight, or where light is poor. A clean visible label is cheaper to serve with print.
Do I need special readers?
Phones work for occasional reads. High volume at a fixed position is more comfortable with a dedicated reader.
Can a tag record when an item moves?
No. A tag is passive and reports nothing, so every record comes from someone arriving and tapping.