Monthly data note — September 2026
September in data: URL accounts for over half of QR Code generations
September recorded 37,773 generations at QR Plus. URL gained share with nearly stable volume; Pix kept a share close to August. The analysis describes the month's composition and generation activity.
September's picture: destinations and generation activity
Between September 1 and 30, QR Plus records show which QR Code types were generated and how activity was distributed across the week.
This edition examines generations concentrated in URL, Pix and WhatsApp, the difference between share and count, and weekly activity. August provides a reference for composition. Platform records describe observed actions without identifying the reason for each creation or the outcome after scanning.
The analysis covers the complete month, September 1–30, 2026. Data was extracted on October 2, with all 30 days covered, no pending events from the period in the queue, and reconciliation between facts and daily aggregates. The monthly file follows the system schedule starting on day 3; it had not been generated at extraction time.
Read the August 2026 studyURL exceeds half of observed September generations
In September, QR Plus observed 37,773 eligible generation events. URL represented 53.03% of the total. This describes the destination type selected when creating a code, before any scan, conversation or purchase.
September and August: counts and shares side by side
| Destination | August: generations | August: share | September: generations | September: share |
|---|---|---|---|---|
| URL | 20,198 | 46.08% | 20,031 | 53.03% |
| Static Pix | 12,889 | 29.40% | 11,063 | 29.29% |
| 6,719 | 15.33% | 2,871 | 7.60% | |
| Other categories | 4,029 | 9.19% | 3,808 | 10.08% |
| Total across all categories | 43,835 | 100.00% | 37,773 | 100.00% |
URL, static Pix and WhatsApp totaled 33,965 generations, approximately 89.92% of September. These destination categories distinguish tasks: informing, receiving and talking. Classification does not reveal URL page content or the user's economic sector, and cannot identify how many codes were offers, stores or commercial campaigns.
A larger share does not mean more generations
URL moved from 46.08% in August to 53.03% in September, approximately 6.95 percentage points. Its monthly count moved from 20,198 to 20,031, a change of −0.83%. Share rose because other categories fell more. August had 31 days and September 30; this descriptive comparison does not demonstrate growth in demand, adoption or business performance.
Pix went from 12,889 to 11,063 events and WhatsApp from 6,719 to 2,871. The totals do not identify why they changed: user acquisition, interface and instrumentation can affect what is observed. Without investigating those factors, the decline should not be interpreted as rejection of payments or messaging-based service.
The base describes QR Plus operational events, not a representative sample of Brazilian retail. Generations are not unique people, scans, orders or sales. The comparison uses the published August study and complete September data. Read the August 2026 study
The October 2 extraction covers the full month and was reconciled between events and aggregates. The monthly file is a separate product, generated starting October 3. No September monthly file is associated with this version; the methodology records the extraction used for subsequent verification.
September's rhythm: higher daily averages Monday through Friday
September recorded 32,427 generations across 22 Mondays through Fridays and 5,346 across eight Saturdays and Sundays. Daily averages were 1,474.0 and 668.3 respectively: 54.7% lower on weekends. Averages account for different numbers of days in each group.
The classification follows the weekly calendar, including holidays falling Monday through Friday. It describes when codes were generated, not when they were scanned or used. It cannot attribute the difference to working hours, retail traffic or consumer preferences. Two months are also insufficient to establish seasonality.
What September's categories help explain
URL: the offer must continue at the destination
Link to a campaign page so customers can find the offer directly. A dynamic URL QR Code allows the destination to be updated; this does not replace checking prices, deadlines and availability on the page.
URL QR CodeWhatsApp: opening the conversation is only the start
Identify the campaign in the initial message and state service hours. Before expanding distribution, review who responds, what details they request and how pickup or delivery is confirmed.
WhatsApp QR CodePix: payment requires a check
The generator’s Pix QR Code is static. Check payment details and ask customers to verify the recipient in their banking app. Do not attribute dynamic destination editing or scan metrics to this code.
Static Pix QR CodeWhat to observe at each step
Following the journey avoids confusing interest with revenue. Use records your business actually has and compare the same period, without attributing information to the QR Code that it does not measure.
| Step | Possible record | Useful question |
|---|---|---|
| Access | Destination visits where supported, and campaign page records. | Did the page open and show the expected offer? |
| Conversation | Campaign-identified conversations and time to the first response, recorded by customer service. | Did the customer receive advice while the offer was available? |
| Order | Orders attributed to the campaign in the store’s system or records. | Could inquiries, reservations and confirmed purchases be distinguished? |
| Payment | Payment confirmed by the business and linked to the order. | Was the order paid, delivered, canceled or returned? |
If you report a rate, state which step is in the numerator and which is in the denominator. Orders per conversation and payments per order answer different questions. Without a link between records, report the steps separately rather than inventing a conversion attributed to the QR Code.
Before printing and promoting
- Open the destination on different phones and check loading, price, deadline and campaign terms.
- Scan a sample at its final size and on its final material, under realistic lighting and distance.
- Prepare responses for questions, stock shortages and ended offers; review the destination when the campaign ends.
- Distinguish code generation, scanning, conversations, orders and payments when evaluating results.
A generated QR Code is not a scan or a sale. Access metrics depend on the QR Code type and available feature. Conversion requires measurement at the destination or in the sales operation.
Printed material keeps circulating after the offer ends
Plan what happens to the destination when a campaign ends. A dynamic URL destination can be changed to announce the end of a promotion and present current alternatives. Static codes retain the encoded information; maintenance depends on content at the original address or replacing the material. Check content, terms and response capacity before reusing a piece.
A question for the story
Why did URL exceed half of generations without increasing its monthly count? How can the mix of types be explained without confusing share with demand? How was activity distributed across the week? September's picture supports these questions for interviews about QR Code use and measurement.
A brief outlook for October
September's data can guide questions for the next month: does the destination work well on mobile, can staff respond, and do records distinguish creation, access and outcome? October will be covered in its own edition with its data and context. This analysis stays centered on September's observed actions.
Sources and scope of this analysis
- Sebrae RS — 2026 retail dates and planning
- Sebrae SP news agency — final four-month sales season (August 25, 2026)
The quantitative base is QR Plus's October 2 extraction covering all 30 days of September, compared with the published August study. The links below provide commercial planning context; they do not explain observed changes or demonstrate campaign performance.