What do QR vendors’ numbers really tell us?

Imagine a restaurant that signed up with a QR generator in 2020. In 2026, it creates a new code for its terrace. The account is six years old; the QR code on the poster is new.

That distinction is central to our reading of Bitly and QRCodeKIT’s reports. Their figures shed light on activity within their platforms. To assess their conclusions, we need to examine the population, period and measurement behind each claim.

63% of scans: account age or code age?

QRCodeKIT reports that 63% of scans recorded from September 1 to 25, 2026 came from accounts created at least five years earlier. It divides accounts into cohorts: groups defined here by their creation year.

Its second trend concludes that most scans come from codes set up years ago. Yet account age does not establish code age. Our restaurant would count as an older account even with its new code. Another might still use its 2020 code: the measurement cannot distinguish these cases.

Establishing how long a printed QR code remains in use would require tracking it over time: its creation, its introduction into circulation and its observed activity. The vendor’s conclusion may be plausible, but this figure does not demonstrate it.

The definition of “still scanned” also matters: an account qualifies if its codes received at least one scan during the twenty-five days observed. That establishes recent activity, not continuous use since account creation.

Do the reports measure the same thing?

Bitly’s 1.9 million creations are a flow: new codes produced during six months. QRCodeKIT’s 2.1 million codes are a stock held by the accounts studied, accumulated over time. Scans measure something else again: activity during a period.

Adding these volumes cannot produce a worldwide total. The populations and periods differ; one business may also use several generators. The reports offer complementary views of creation and use, but do not measure the same population with the same indicator.

What does 6.2% scan growth establish?

Bitly reports 6.2% scan growth on its platforms between the first halves of 2025 and 2026. This is a change in scan volume relative to the starting volume. It measures neither growth across the whole QR market nor a conversion rate.

New customers, more codes distributed or one large campaign can raise the total. A constant account base, following the same accounts over both periods, would separate out growth from new accounts. Code counts and campaigns would still need examining. The total alone cannot distinguish these effects.

A conversion rate measures the share of visitors who take an intended action, such as ordering or signing up. It requires tracking that action and the relevant visitors. Scan counts, which can include repeat scans by one person, do not provide that rate.

What is the denominator, and who is in the sample?

The denominator is the total used to calculate a percentage. For the 63% figure, it is all scans observed during the period. This is therefore not the proportion of older accounts that remain active.

One customer can create many codes, with a few receiving most scans. The unit determines how much weight each use carries:

• Accounts: each account counts once.
• Codes: an account holding many codes carries more weight.
• Scans: heavily scanned codes dominate the result.

The sample, the accounts, codes or scans included in each analysis, also limits how far the findings can be generalised. An agency-heavy platform may see many campaigns; a restaurant-heavy one, repeated visits. A large base does not guarantee market representativeness. Generalising requires examining account selection and the mix of industries, countries and uses.

How should you cite and use these studies?

Keep the platform, period, population and measured indicator with each figure. “On QRCodeKIT, 63% of scans observed from September 1 to 25, 2026 came from accounts created at least five years earlier” preserves the finding; inferring printed code age requires further data.

These reports provide useful observations and hypotheses to investigate. Compare your campaign with similar placements, audiences and periods, then measure its intended outcome. When choosing a tool, check its features and terms: our comparison methodology explains the criteria, and the QRBench statistics page separates our measurements from third-party data.