A QR code generator with analytics can only count scans when the code is dynamic. A dynamic code holds a short redirect link, and the scan is logged when that link forwards the phone to your real page. A static code points straight at the destination, so nothing can count it. The dashboard is not the question that matters most. What matters is who controls the redirect domain, because if it stops working, every printed code fails with it.
TL;DR:
- Scan tracking only works on dynamic QR codes. The analytics are server logs from the redirect behind the code, not something stored in the image.
- Total scans, time, device type and country are reliable. "Unique scans" and city-level location are estimates.
- Many free plans shut off dynamic codes after a trial, which turns your printed codes into upgrade prompts.
- Use 302 redirects, not 301. Phones can cache a 301, so repeat scans never reach your server and never get counted.
- If you print codes at volume, owning the redirect domain matters more than any dashboard feature.
Table of Contents
- Static vs. Dynamic Codes: Where QR Analytics Come From
- What a QR Code Generator With Analytics Measures, and What It Only Estimates
- The Free-Tier Trap: When "Free" Analytics Break the Code You Printed
- Features Worth Paying For, and the Ones That Are Padding
- Styling and Logos Without Losing Scans
- Your Scan Logs Are Personal Data
- Hosted, Self-Hosted, or Plain UTM Tags: Which Would We Pick?
- Own the Domain Behind Every Code You Print
- Frequently Asked Questions
- Sources
Static vs. Dynamic Codes: Where QR Analytics Come From

No QR code reports anything by itself. The image is a pattern of modules that encodes a string of text, as defined in the ISO/IEC 18004 standard. When that string is your final URL (a "static" code), the phone goes straight to your site. No third party sees the scan, so nothing can count it except your own website analytics.
A "dynamic" code encodes a short link on a redirect server instead. The phone requests that short link, and the server records the request: timestamp, IP address, user agent and referrer. Then it sends the phone on to your real destination. Every analytics dashboard you have ever seen is a summary of those redirect logs.
This gives dynamic codes two things static codes cannot have:
- Scan counts you don't have to set up per page. The redirect logs every hit, whether or not the destination page runs any tracking script.
- An editable destination. You can change where the code points after 10,000 flyers are printed, because the image only contains the short link.
The cost is dependency. A dynamic code only works while the redirect server answers. That one fact should drive your choice of generator more than any feature list.
What a QR Code Generator With Analytics Measures, and What It Only Estimates

Most dashboards show the same dozen metrics. Some of them are measured directly. Others are inferred, and inferred numbers deserve less trust. Here is how each one is actually produced.
| Metric | How it's captured | How much to trust it |
|---|---|---|
| Total scans | One server request per redirect | High, once bots are filtered |
| Time and date | Server timestamp | High |
| Device / OS | Parsed from the user-agent string | High for OS, moderate for exact model |
| Country | IP geolocation lookup | High |
| City | IP geolocation lookup | Moderate to low. Mobile carriers often route traffic through distant gateways |
| Unique scans | Guessed from IP + user agent, or a cookie | Low. Many phones share one carrier IP |
| Conversions | Needs tracking on the destination page | Only as good as your destination tracking |
Two things distort these numbers more than most buyers expect.
- Bots and link previews. Messaging apps fetch a preview when someone shares the short link, and email security scanners open links to check them. Each fetch looks like a scan unless the generator filters known bot user agents.
- Shared mobile IPs. Carriers put large numbers of phones behind a small pool of public IP addresses. "Unique" counts based on IP can merge many different people into one visitor.
Pro Tip: Check which HTTP status your generator uses for redirects. Under RFC 9110, the IETF's HTTP semantics standard, a 301 response is cacheable by default and a 302 is not. If your codes return 301, a phone's browser may remember the destination and skip your server on the next scan. That undercounts repeat visitors, and on that device, later destination changes can fail to take effect. Run curl -I on your short link. If you see 301, ask the vendor to switch to 302 or 307.
A 301 redirect can be cached by the phone. The second scan never reaches your server, so it never shows up in your analytics.
The Free-Tier Trap: When "Free" Analytics Break the Code You Printed
Search for a free QR code generator with analytics and you will find plenty. Read the terms before you print. A common pattern works like this:
- A free trial creates dynamic codes with full tracking.
- The trial ends, or you pass a scan limit.
- The codes start redirecting to a page that asks you to upgrade, or they stop resolving.
The code itself never changes. The vendor simply changes what its server does when your short link is requested. Because that short link sits on the vendor's domain, you cannot point it somewhere else. Your only options are to pay or to reprint.
You can find out whether a free plan works this way before you commit. Look for three answers in the pricing page or terms:
- Does the code keep resolving if you stop paying? You need a plain yes. "Your data is retained" is a different promise.
- Can you use your own domain? If the short link is
go.yourbrand.com, you control where it points no matter which tool sits behind it. - Is there a scan cap? Some free tiers stop counting, or stop redirecting, after a fixed number of scans per month.
If the answer to the first question is unclear, treat the code as a rental.
Features Worth Paying For, and the Ones That Are Padding
Comparison pages tend to list 30 features at equal weight. In practice, five of them decide whether the tool holds up after a year of printed material.
Custom domain on the short link
This is the most important feature, and it serves two purposes. It lets you move to another tool later without reprinting. It also shows people a domain they recognize. The FTC warns consumers that scammers hide harmful links in QR codes and tells them to check the URL before opening it. A link on your own branded domain passes that check. A random string on an unfamiliar domain may not.
UTM pass-through to your own analytics
The redirect should append campaign parameters so the visit shows up correctly in your web analytics. Google Analytics documents the standard parameters: utm_source, utm_medium and utm_campaign. Use one code per placement (poster, packaging, table tent) so each placement gets its own source. One code reused across every placement tells you almost nothing.
Raw data export
You want CSV or API access to scan-level rows, not just charts. Without it, your history lives only inside the vendor's dashboard.
Bulk generation
If you print serialized labels, event badges or per-location signage, you need to create hundreds of codes from a spreadsheet, each with its own destination and tag. Making them one at a time does not scale past a dozen.
A/B destinations and access control
Splitting traffic from one printed code between two landing pages turns a poster into a real experiment. Our split testing tool guide covers what a fair split requires. For gated content such as internal menus, price sheets or event materials, password-protecting the destination URL is often simpler than building a login.
Padding to ignore: "heatmaps" of scan locations at street level (IP data is not that precise), AI-generated scan "insights," and separate per-code charges for frame styles.
Styling and Logos Without Losing Scans
Branded codes get scanned more readily, but a logo is damage you add on purpose. QR codes survive damage through error correction, which adds redundant data. Denso Wave, the company that invented the QR code, defines four levels:
- Level L: restores roughly 7% of the codewords
- Level M: roughly 15%
- Level Q: roughly 25%
- Level H: roughly 30%
Level H restores about 30% of a damaged code. A centered logo uses up part of that budget before the code ever gets scuffed, folded or rained on.
Use Level H for any code with a logo. Keep the logo well under that 30% ceiling, around a fifth of the code's area at most, so some margin is left for real-world wear. Dynamic codes also help here. A short redirect URL encodes far fewer characters than a long campaign URL, so the code needs fewer modules, and each module prints larger and scans more reliably. Our guides to creating a branded QR code that won't break and styling codes without breaking the scan cover contrast, quiet zones and color rules.
Two quick notes on related searches. A "Unicode QR code generator" usually means one of two things. Either it encodes non-Latin text, which QR byte mode handles through UTF-8, or it draws the code with Unicode block characters for terminals. Neither one adds tracking. A barcode generator makes 1D codes for inventory and retail scanners. Phones do not reliably open URLs from those, so they are the wrong tool for scan analytics.
Your Scan Logs Are Personal Data
Every scan log contains an IP address, and under EU law that is not anonymous. Recital 30 of the GDPR names IP addresses as online identifiers that can be used to profile and identify people. If anyone in the EU scans your code, the redirect server is processing personal data on your behalf.
With a hosted generator, that means a third party holds your customers' scan records. You need a data processing agreement with them, a clear answer on where the data is stored, and confidence that they are not combining your scan data with data from other customers. With a self-hosted redirect, the logs stay on your server and under your retention policy. That is one fewer processor to list in your privacy notice.
Whichever route you choose, a few habits reduce your exposure:
- Truncate IPs after the geolocation lookup. Country-level reporting does not need the full address stored.
- Set a retention period. Raw logs older than 12–24 months rarely answer a marketing question.
- Don't fingerprint scanners. Cookie-free "unique" estimates are less accurate, and that tradeoff is acceptable.
Hosted, Self-Hosted, or Plain UTM Tags: Which Would We Pick?

There are three workable setups. The right one depends mostly on how many codes you print and how long they stay in circulation.
| Setup | Upfront cost | Ongoing cost | Who controls the link | Best for |
|---|---|---|---|---|
| Static code + UTM tags | Free | None | You | A few codes, destinations that never change |
| Hosted dynamic generator | Low or free | Monthly subscription, often scaled by code count | The vendor (unless you use a custom domain) | Teams with no server access |
| Self-hosted redirect + QR | Setup time or one-time license | Server, roughly a few dollars a month | You | Long-lived printed codes, privacy-sensitive data |
Our position is this. If a code will be printed on anything that outlives a single campaign, such as packaging, signage or business cards, do not let a subscription control it. A hosted plan at $15 a month costs $900 over five years. You keep paying to protect what you have already printed, not to get new features. If you only need three codes for a two-week event, static codes with UTM tags and your existing web analytics are free and good enough.
Hosted tools are the right call when nobody on your team can maintain a server. Be honest about that. A self-hosted redirect that goes down at 2 a.m. breaks every code at once. If you are weighing the build-or-buy question, our breakdown of the self-hosted QR code generator route covers the operational side.
Own the Domain Behind Every Code You Print
Kortli is a self-hosted link shortener that you install on your own Linux server for a one-time license fee instead of a monthly subscription. It produces styled QR codes with your logo, each backed by a short link on your own branded domain. You can add as many custom domains as you like with no per-domain charge. Scan analytics, including device, time and geographic location, stay on your server. If you ever stop using the software, the domain is still yours, so you can redirect it elsewhere and your printed codes keep working. A/B split testing and password-protected links are included.
Check the current license price on the Kortli site. The honest caveat: self-hosting means uptime is your responsibility. You need a server you keep patched and backed up, plus basic monitoring. City-level location is subject to the same IP-geolocation limits described above, as with every tool in this category. Read our spec sheet the same way you'd read anyone else's.
Frequently Asked Questions
Can a static QR code track scans?
No. A static code sends the phone directly to its destination, so no redirect server sees the scan. You can only measure the visit afterward, using UTM parameters and the analytics on your destination page.
Is there a truly free QR code generator with analytics?
Some exist, but many free plans deactivate dynamic codes or redirect them to an upgrade page once a trial or scan limit runs out. Before printing, confirm in writing that codes keep resolving if you stop paying. Better still, use a generator that supports your own domain.
How accurate is QR code scan location data?
Country-level location from IP geolocation is generally reliable. City-level data is much less reliable, because mobile carriers often route traffic through gateways far from the person scanning. Use city data as a rough signal, not a fact.
Why do my scan counts look higher than my website visits?
Link-preview bots in messaging apps and email security scanners fetch short links, and some dashboards count those fetches as scans. People also scan and then close the page before it loads. A gap of some size between scans and sessions is normal.
What error correction level should a QR code with a logo use?
Use Level H, which can restore roughly 30% of the code. Keep the logo well below that limit so some correction capacity is left for real-world wear, and test-scan the printed proof on several phones.
Do I need a privacy policy update to track QR scans?
If people in the EU scan your codes, almost certainly yes. Scan logs contain IP addresses, which the GDPR treats as personal data. Your notice should say what you collect, who processes it and how long you keep it.
Sources
- Error Correction Feature | DENSO WAVE (QRcode.com)
- RFC 9110: HTTP Semantics | IETF / RFC Editor
- Scammers hide harmful links in QR codes to steal your information | Federal Trade Commission
- Recital 30: Online Identifiers for Profiling and Identification | GDPR-Info.eu
- Collect campaign data with custom URLs | Google Analytics Help

Leave a Reply