How many items can one person scan per hour with NFC?
The tap takes about a second and reaching the item takes far longer, so the figure is set by your building rather than by the technology. Measure a real aisle with a real operator before promising a number, because no published figure describes your site.
Walking dominates
A read is roughly a second, including presenting the phone and getting a confirmation.
Reaching the item, finding the tag on it, and moving to the next one is where the time goes. In a warehouse that is often measured in tens of seconds per item.
So the technology is not the bottleneck, and a faster reader improves a fraction of a fraction.
This also means NFC and a printed code produce similar throughput for the same walk, which removes speed as an argument between them.
Measure your own
Take one operator, one real aisle, and a stopwatch. Count items completed in twenty minutes and multiply.
Do it with the equipment they will actually use, wearing what they will actually wear, at the time of day the work happens.
Then do it again on a different aisle, because a dense shelf and a scattered yard produce very different figures and the average of the two describes neither.
Twenty minutes of measurement is worth more than any published benchmark, and it gives you a number you can defend.
What actually speeds it up
Tag position consistency. If every tag is in the same place on every item, the operator stops searching, and searching is most of the per item time.
A marked target on the tag, so the phone goes to the right spot first time.
Route order. Scanning in the order items are arranged beats scanning in the order a list was printed.
And a device that confirms clearly. An operator who is not sure a read succeeded taps again, which doubles the cost of every uncertain read.
What to promise
A measured figure from your own site, with the conditions stated: which aisle, which operator, which equipment.
Not a figure from a supplier, which describes a demonstration rather than your building.
Include the failure rate in the figure. Items that need a second attempt, or a walk back, are part of the real cadence.
And revisit it after a month of real use, because the first week is learning and the figure improves once operators stop searching for tags.
Frequently asked questions
Is NFC faster than a barcode for stocktaking?
Not meaningfully, since walking dominates both. NFC wins on dirty labels and poor light rather than on speed.
Would a dedicated reader be quicker?
Slightly per read, and the walk is unchanged. It helps most where the items come to a fixed station.
Why do my operators tap twice?
Usually because the confirmation is unclear. An operator who is not sure a read succeeded repeats it, which is pure cost.
Can I use a supplier's throughput figure?
Only as a demonstration. Measure a real aisle with a real operator, because the building sets the number rather than the technology.