How do you measure whether a card is used?
You measure arrivals at your page, never taps. The card is passive and reports nothing to anyone, so everything you can possibly know comes from your own server. That limit is worth understanding before a supplier promises you tap analytics.
The card measures nothing
A card has no battery, no clock and no memory of what has happened to it. It answers when powered and forgets immediately.
Some chips carry a counter that increments on reading, which is the nearest thing to on card measurement and requires reading the card back to consult. That is useful for a tag you own and useless for a card you gave away.
So nothing about usage comes from the card. It all comes from what happens after the tap.
What you can actually see
Requests arriving at your page. That is the entirety of the raw material.
If each card resolves to its own address, or carries a mirrored identifier, you can tell which card produced which arrival. That is what makes per person measurement possible in a team deployment.
What happens on the page afterwards is measurable by the usual means, subject to whatever you choose to record and whatever you tell visitors about it.
And nothing at all if the card points at a third party platform, since the arrivals are theirs rather than yours.
The gap between taps and arrivals
Not every tap becomes an arrival, and the difference is not noise.
An iPhone shows a notification that has to be tapped, and so does Android for a web link from Android 17. A dismissed notification is a successful read that produced no arrival at all.
A phone with no coverage reads the card and fails to reach the page. A person who taps twice by accident produces two arrivals from one intention.
So arrivals undercount interest in some ways and overcount it in others, and treating them as a tap count is wrong in both directions.
Measuring something useful
Count the action rather than the visit. If the point of the card is that people save your contact or book a call, that is the number that matters, and it is the one a page can record.
Compare periods rather than chasing absolute figures. A card handed out at an event produces a shape over the following days that tells you more than a total.
Ask people. In a small team, the question of whether the cards are working is answered faster by asking than by instrumenting.
And decide what you will do with the number before collecting it. Measurement that changes no decision is a cost with no return, and it carries obligations towards the people it counts.
Frequently asked questions
Can I count how many times my card was tapped?
No. The card reports nothing, so you can only count arrivals at the page, which is a different and imperfect proxy.
What is a read counter?
A chip feature that increments when the card is read. It is consultable only by reading the card back, so it suits tags you keep rather than cards you give away.
Do I see anything if I use a card platform?
Whatever the platform shows you, and the arrivals belong to them. Using your own domain keeps the measurement yours.
Why do arrivals seem low?
Dismissed notifications, taps with no coverage and people who simply do not follow through all produce reads with no arrival.