# What scan rate should you expect?

> Nobody can tell you an average QR scan rate, because scan rate is scans divided by impressions and no platform can count who walked past a poster. Published vendor averages are survivorship-biased marketing. What actually predicts scans, in order: dwell time, a specific stated benefit, size for distance, and destination quality.

Source: https://useqr.app/docs/marketing/what-scan-rate-should-you-expect · Last reviewed 2026-08-21 · UseQR is free forever, no signup.

---

## The denominator problem

[Scan rate](/glossary/scan-rate) is a fraction: scans ÷ impressions. The numerator is easy: [tracking scans](/docs/how-to/how-to-track-qr-code-scans) is a solved problem. The denominator is the whole difficulty: **nobody counts how many people saw a poster, held a package, or glanced at a flyer.** Print has no impression log. Any percentage quoted without a stated, measured denominator is not a rate; it is a number wearing one's clothes.

This is not pedantry: it is why the question "what is a normal scan rate?" has no honest general answer, and why anyone offering one confidently is selling something.

## Why vendor "average scan rates" mislead

QR platform marketing pages publish average scan rates, and the figures are structurally unreliable in three stacked ways:

1. **The denominator is invented.** The platform observes scans on its codes. It observes nothing about impressions. Those are estimated, assumed, or quietly replaced by "codes created" or some other countable stand-in.
2. **Survivorship bias, twice.** The customers of a paid QR analytics platform are self-selected, engaged marketers running above-average campaigns, hobby codes and failed campaigns are underrepresented from the start. Then campaigns that perform badly churn, removing themselves from every future "average".
3. **Averaging across channels is meaningless.** A menu code scanned by half of diners and a billboard scanned by nearly nobody do not have a useful mean, any more than restaurants and billboards have an average temperature.

None of this requires bad faith. It falls out of what a platform can and cannot see. But it means published ranges vary wildly and none transfers to your campaign.

## What actually predicts scans, ranked

The variables below are worth more than any benchmark, in roughly this order of effect:

1. **Dwell time.** A seated, waiting, or queueing viewer scans; a walking one does not. This single variable separates [menus and transit panels from billboards](/docs/marketing/qr-codes-in-out-of-home-advertising).
2. **A specific stated benefit.** "See tonight's specials" versus a bare code, the [call-to-action page](/docs/marketing/qr-code-call-to-action) ranks this second only to dwell.
3. **Size for distance**, at least a tenth of the viewing distance, or the decision never even occurs.
4. **Reachability**: eye level, unobstructed, physically approachable.
5. **Destination quality.** A slow or bait-and-switch landing page does not just lose that visitor; it teaches people not to scan your next code.

## Honest expectations by channel

Qualitative, deliberately, expressed as the share of people who *handle or sit with* the medium:

| Channel | Expectation |
|---|---|
| Menus, table tents, in-hand packaging | a meaningful minority, the best case in print |
| Direct mail, at home | small but real, and precisely measurable per piece |
| Counter and till cards | small minority, repeat exposure helps |
| Street posters, walk-by | a small fraction of a small fraction |
| Roadside billboards | approaching zero |
| TV | a tiny fraction of a huge base: absolute scans can still be large |

## Getting a number you can actually use

Replace the unknowable denominator with one you control: **scans per 1,000 pieces mailed**, per event attendee, per menu per week, per location per day. Those are real, comparable, decision-grade metrics, and comparing your own placements against each other, via [A/B testing with per-placement parameters](/docs/marketing/ab-testing-qr-code-placement), answers the question you actually have, which was never "what is average?" but "which of my options is best?". Whatever you deploy, [verify it decodes](/validate) first: an unscannable code has a rate of exactly zero, and nobody phones to tell you.

## FAQ

### What is the average QR code scan rate?
There is no reliable average. Scan rate requires counting impressions, and impressions of print are uncounted and uncountable. Published vendor averages rest on invented denominators and survivorship-biased samples, which is why their figures disagree so widely.

### What is a good scan rate for a QR campaign?
One better than your own last campaign. Define a denominator you control (pieces mailed, attendees, menus per week) measure scans against it, and improve placement, wording and size iteratively. External benchmarks cannot survive the transfer to your context.

### What makes people actually scan a QR code?
In order: being stationary with time to spare, a specific stated benefit, a code large enough for the distance, easy physical reach, and trust that the destination is worth it. Design controls all five; luck controls the rest.

### Do more people scan QR codes now than before?
Adoption rose sharply when phone cameras gained native scanning and hospitality moved menus to codes: scanning is now an unremarkable habit in most markets. That raises the ceiling, but placement and incentive still decide whether any individual code gets scanned.

## Try it

- https://useqr.app/url
- https://useqr.app/validate