Opens in a new tab

How QR code tracking works, and what the numbers don’t tell you

by: Wojciech Filipek
·
·
November 18, 2026
Share
IN THIS ARTICLE

A QR code can’t track anything by itself. It’s a printed pattern, and the phone that reads it doesn’t report the scan to you. What gets tracked is the visit that follows: a tracked code points to a short link on a server, which records the request and sends the person on to the real page. Strictly speaking, what you count is people who opened the link, not everyone who pointed a camera at the code.

So yes, you can track QR code scans, with a dynamic code or with analytics on the page it opens. Codes that hold Wi-Fi details, a contact card or plain text open nothing on the web, so they can’t be tracked at all. We run Scanteo, our own dynamic QR code platform, and most of what follows comes from building and running its scan analytics.

What happens during a tracked scan

  1. The phone camera reads the code and finds a short link, something like qr.example.com/a7Kx.
  2. The person taps it and the phone requests that link from the QR platform’s server.
  3. The server logs the request: the time, the code, the device and browser it reports, and an approximate location worked out from the IP address.
  4. The server redirects the phone to the destination you’ve set: a menu, a product page, a booking form.

The whole detour takes milliseconds. On Scanteo’s redirect service we load-tested 2,000 requests a second with a p95 response time of 6.8 ms in a test environment, which is far more than any printed campaign sends. A slow redirect is the one thing a QR platform must never have, because people blame your brand, not the platform.

A static code skips steps 2 to 4. The destination is written into the pattern itself, so the phone goes straight there and nobody in between sees the scan. That’s why static codes can’t be tracked, and also why they never expire. The QR code guide covers the difference in more detail.

What a scan record can tell you

A good scan report answers four questions:

  • How many scans, and when. By day and by hour, which tells you whether the lunchtime table card or the evening poster does the work.
  • Which code. If each placement has its own code, you can compare them directly.
  • Which device. iPhone or Android, sometimes the browser.
  • Roughly where. Country and usually city.

That’s enough to make real decisions. It’s not enough for everything people expect from it.

What the numbers don’t tell you

Scans aren’t people. If someone scans the same menu code three times during dinner, that’s three scans. Platforms estimate “unique scans” by grouping requests from the same device over a period, but it’s an estimate. Treat the trend as reliable and the absolute number as approximate.

Location is a guess. It comes from the IP address, and a phone on mobile data can show up in a city 50 miles away. A VPN moves it to another country. Country-level data is solid; street-level data doesn’t exist.

Some hits aren’t scans. Link-preview bots and security scanners open links too, especially when a code’s link is shared in a message or email. A decent platform filters known bots. If your numbers jump overnight with no campaign behind them, check this first.

A scan isn’t a result. The report tells you someone arrived, not whether they booked, ordered or left after two seconds. For that you need the destination side: UTM parameters on the destination link, so your website analytics shows the visit as coming from a specific code and placement.

Cookie-free tracking and why it matters

Scan tracking on the redirect doesn’t need cookies. The server sees one request and logs it; nothing is stored on the phone. That’s how we built Scanteo’s analytics: scans are counted without cookies, and the platform is hosted in the EU.

This matters if you have visitors from Europe. The moment your tracking stores something on the phone, you’re in consent-banner territory, and a large share of people decline. Cookie-free counting on the redirect gives you complete scan numbers; website analytics behind a consent banner gives you partial ones. Use both, and expect them to disagree.

How to set up QR code tracking you can trust

  1. Use a dynamic code for anything printed in volume or kept around for months. Tracking is a bonus; the real reason is that you can change the destination after printing.
  2. Make one code per placement. A restaurant with table cards and flyers should print two codes, not one, or the report can’t tell them apart. Assume 20 table cards and 500 flyers: one code for the cards, one for the flyers, and next month you know which deserves the print budget.
  3. Add UTM parameters to the destination, for example ?utm_source=qr&utm_medium=print&utm_campaign=table-card, so the visit shows up correctly in your website analytics.
  4. Check what happens when you stop paying. Some tools switch codes off when a subscription ends, and every printed code stops working with them. Scanteo keeps printed codes working after a downgrade or cancellation, and we’d look for the same in any provider.
  5. Watch the destination as closely as the scans. A code that scans perfectly and opens a 404 page is worse than no code. Scanteo checks destination links daily for this reason.

Free QR code tracking: what to expect

Most “free QR code tracking” offers are trials of dynamic codes. That’s fine for a test, but read the terms before you print: if the free plan ends and the codes stop redirecting, you’ve paid for the print run twice.

A free alternative that never expires: a static code pointing to a URL with UTM parameters. You lose scan counting on the redirect and the ability to change the destination, but your website analytics still shows visits from that code.

When a QR tool isn’t enough

Off-the-shelf QR platforms cover almost every marketing use. The exceptions are codes that are part of a product: a code on every unit in a warehouse, codes that open a customer’s own account, or scan data that has to land in your CRM or ERP in real time. Those need an API, webhooks or a system built around your process, and that’s the kind of custom software we build. For a menu, a poster or packaging, pick a dynamic QR tool, one code per placement, and check the destination before you print.

Frequently asked questions

Only roughly. The location comes from the IP address of the phone that opened the link, which is reliable at country level and usually close at city level. A QR code can’t read a phone’s GPS. Only the page it opens can ask for that, and the person has to allow it.

Scanning a code doesn’t give the code owner your name, phone number or contacts. The tracking server sees what any website sees: the time, your device and browser, and an IP address. The real risk is where the code leads, such as a fake login or payment page, so check the address before you enter anything.

Yes. Point a static code at a URL with UTM parameters and your website analytics will show visits from that code. You won’t get scan counts on a redirect or the option to change the destination later. Free dynamic-code trials work too, but check what happens to the codes when the trial ends.

A dynamic QR platform shows total scans and an estimate of unique scans, grouped by device over a period. Treat the unique number as an estimate. For a clean comparison, print a separate code for each placement.

Not with scan tracking alone. Scan counts are anonymous, so they tell you how many devices opened the link, not who they belong to. Attendance needs a check-in step that identifies the person, such as a login, a ticket number or a code unique to each guest.