Two separate events get compressed into that one word, and the gap between them explains most of the confusion around QR scan numbers. There is the physical event. A camera decoding a pattern, which no server anywhere can see or count. And there is the tracked event, a click on the link that pattern contained. That second one is the only part any analytics tool actually measures.
The decode and the click are not the same thing
When a phone camera reads a QR code, it decodes the pattern into text, almost always a URL, entirely on the device itself. That step happens with no network connection required. It leaves no record anywhere. If the code encodes a bare destination URL with nothing tracking it, the decode is the whole event. Nobody, including you, will ever know it happened.
What gets reported as a "scan" in any analytics dashboard is actually the next step: the phone's browser opening that decoded URL, which then hits a server that logs the request. If that URL is a tracked redirect or a page with analytics behind it, the hit becomes a data point, a timestamp, a device, a rough location. If it is not, the decode still happened, but nothing downstream of it ever will.
QR scan vs visit: why the two numbers rarely match
A QR scan and a page visit are the same thing in most practical setups, since scanning is simply the method someone used to arrive at a link, no different in principle from tapping a link in a text message. But three things routinely make the reported scan count diverge from the true number of people who engaged with the code:
| Event | What triggers it | Does an analytics tool see it | Counted as |
|---|---|---|---|
| Camera decode | Any phone camera reading the pattern | No, happens entirely on-device | Not counted at all unless it leads to a link open |
| Scan / link open | The decoded URL loading in a browser | Yes, if the URL is tracked | A tracked scan or click |
| Preview fetch | A messaging app pre-loading the link to build a preview card | Yes, often indistinguishable from a real open | Frequently miscounted as a real scan |
| Unique scan | A first-seen device or cookie hitting the tracked link | Yes, deduplicated by the tracking tool | One entry per distinct device, not per distinct person |
How accurate is QR scan tracking, honestly
Reasonably accurate for what it actually measures, and consistently misread for what people assume it measures. A tracked QR code will reliably tell you that a link was opened, roughly when, on what kind of device, and from around what location. It will not tell you, with any precision, how many distinct human beings that represents, because of one specific gap most tools do not surface clearly: a screenshot of a QR code scanned by five different people on five different phones produces five legitimate, correctly counted scans. But a single QR code sitting on a shared surface, a reception desk tablet, a kiosk, a shared print copy passed hand to hand, can generate ten scans from what your dashboard will happily record as a handful of "unique" devices, because the tool is deduplicating by device fingerprint, not by person. Unique scan counts are a device-level proxy for people, and that proxy quietly breaks in any shared-device or repeat-scanning scenario, which is common at events, in retail, and on printed material that gets passed around.
What counts toward the scan number and what does not
A real link open counts every time, whether it came from the original printed code or from a photo of it, and whether someone scanned it once or scanned it five times. A messaging app or browser fetching the link to build a preview thumbnail usually counts too, and that is where the inflation creeps in, since nobody has actually looked at anything yet. The camera decode itself never counts on its own, not on any tool that only tracks the resulting link open rather than the decode event, which describes nearly all of them. And a repeat scan from the same device does not register as a second person, at least not once the tool is reporting unique scans rather than total scans.
Total scans vs unique scans
Most tools report both numbers. They answer different questions. Total scans is every link open the tracked code produced, including the same person scanning it three times over a week. Unique scans deduplicates that down to one entry per distinct device or cookie seen. Neither one is "the real number of people," they are two different approximations of it, and the honest way to use them is comparatively, this placement against that one, rather than as a precise headcount. The concept mirrors unique clicks versus total clicks on an ordinary link, since a QR scan is a click that happened to arrive by camera instead of a tap.
Why a QR code scan is really just a link click in disguise
It is worth saying plainly. It resolves most confusion at once: nothing about the QR image itself is being measured. The pattern is a delivery mechanism for a URL, and once that URL loads, everything from that point forward is ordinary click tracking, the same mechanism behind any tracked link. See what a QR code is for the format itself, and how to track QR code scans for the practical setup that turns a printed code into a source of that data in the first place.
What a scan record cannot tell you
Even a correctly tracked scan gives you a timestamp, a device type, and country and city level location in most tools, not a name, an email address, or confirmation of exactly who physically scanned it. It also cannot see anything that happens after the person leaves the page the code sent them to. If that page then links onward to a third-party site, the scan record ends at your own link, since a QR tracking tool has no visibility into a website it does not serve.