Posted: August 29th, 2011 | Author: admin | Filed under: General | No Comments »
The dominance of iPhones over Androids as the most popular QR code scanning device has continued to increase over the past 6 months based on figures from our own database of just under 1.2 million scan events.
While the Android was the preferred choice prior to February this year, their share has been decreasing steadily since August last year to the point where last month there were 1.9 scans on an iPhone for every one on an Android. Both Blackberry and Windows devices are struggling along with just over 10% and well under 5% respectively.
Considering the number of Android devices out there compared to iPhones, I’ll leave it up to the marketing guys to explain why the numbers stack up the way they do. It could be the demographics of iPhone vs Android users, the relative ease of use (and quality) of the scanning apps available for each platform, or a device-specific bias caused by many placements only showing an iTunes logo in the instructions on how to use the QR code. Who knows?
Just a quick note – unlike may other QR code generation platforms, I don’t force my users to use our URL shortener (a short URL redirection is the standard method for collecting analytics). When given the option of encoding their own website URL directly into the QR code, rather than one of our short URL’s which then redirects to their website address, 80% of users choose the former. As a result our data set only includes analytics data for 20% of the 441,000 QR codes containing website URL’s that QR Stuff users have created over the past 6 months.
Having said that, 1.2 million scan events is still a statistically significant data set – the encoding preferences of the person creating the QR code are a variable that’s completely independent of the phone preference of the person scanning the QR code.
By the way, if you’re curious about why people would choose not to use a URL shortener, see my previous article on Using Google Analytics With QR Codes for a few very good reasons.
UPDATE: Received an interesting email from Eddy Hagen at the Flemish Innovation Center for Graphic Communications with a possible explanation for the large number of iPhone scans showing up in the figures:
We at VIGC have done some testing over the last few months and probably the answer is the following: most (over 75%!) of the QR-codes in magazines, newspapers, brochures can’t be scanned with an average smart phone (e.g. my Nokia E52). Many QR-codes are printed too small, which make them only scannable with smart phones that have a very good camera and macro / zooming capabilities, so the newest iPhones (and alikes). That might be the reason why your figures are not corresponding the number of devices/distribution of devices that are out there.
Posted: August 13th, 2011 | Author: admin | Filed under: General | No Comments »
There’s been a lot of talk lately (and a lot of really cool infographics) about what sort of QR codes are being scanned, where they’re being scanned, and by whom, but these numbers are only part of the story.
Marketers can gain good insights into consumer behavior and demographics from the scan event numbers that are being released, but they’re the effect not the cause. Another important set of numbers is what sort of QR codes are being deployed, and where they are being deployed, and these are necessary to put the scan event numbers into perspective.
Scan event reporting to date has pretty much been about absolute numbers in a vacuum with no relative context brought into play for effectiveness, reach and consumer uptake, apart from anecdotal or survey-style feedback from the consumers themselves. This isn’t a bad thing, it’s just what it is, and makes sense when you consider who the majority of the organizations are that go to the trouble of collating and presenting the data – scanning apps and consumer-facing market research companies – and what data sets they have available.
For example, 20,000 scan events in a particular market is a good number if only 100 QR codes have been deployed into that market, but it represents poor uptake if there’s 100,000 QR codes in the wild in the same market. The relative number of scans achieved for different QR code data types is also not particularly useful if you don’t know the relative number of QR codes deployed for each data type. Social media QR codes accounting for 20% of the reported scan events would clearly be seen as a relatively more engaging type of QR code type if social media QR codes only represented 5% of the total QR codes deployed, but not so much if they were 30% of the deployed total.
Here’s some numbers from the point of view of QR code creation rather than QR code scanning, and relate to the 900,000 QR codes created by the 700,000 user of QRStuff.com over the 6 month period from February to July 2011. Approximately 87% of those QR codes were created anonymously by free users so location data for anonymous QR code creation events has been taken from Google Analytics. In the interests of user privacy, none of the underlying raw or granular data will be (or has been) made available to third parties.
In the tables below I’ve compared the figures from February 2011 and July 2011 and showed relative rankings in each of those 2 months, what the percentages proportions were for the QR codes created in each of those 2 months, and the relative increase in the numbers of QR codes created over the 6 month period from 01 February to 31 July 2011.
Who’s Making QR Codes?
| Position |
Country Of Origin |
Percentage Of Total
QR Codes (Month) |
Increase In Number
Of QR Codes Created
(Feb to Jul) |
| Jul-11 |
Feb-11 |
Jul-11 |
Feb-11 |
| 1 |
1 |
United States |
54.1% |
57.1% |
296% |
| 2 |
2 |
United Kingdom |
12.1% |
7.9% |
475% |
| 3 |
3 |
Canada |
5.4% |
6.0% |
280% |
| 4 |
4 |
Australia |
3.3% |
2.4% |
424% |
| 5 |
5 |
Netherlands |
1.7% |
2.2% |
242% |
| 6 |
10 |
India |
1.6% |
1.1% |
432% |
| 7 |
6 |
Germany |
1.5% |
1.8% |
259% |
| 8 |
34 |
Saudi Arabia |
1.3% |
0.2% |
1689% |
| 9 |
13 |
Mexico |
1.0% |
0.7% |
459% |
| 10 |
7 |
France |
0.9% |
1.5% |
176% |
| 11 |
15 |
Brazil |
0.8% |
0.6% |
411% |
| 12 |
11 |
Denmark |
0.8% |
0.9% |
292% |
| 13 |
27 |
Malaysia |
0.8% |
0.4% |
651% |
| 14 |
9 |
Italy |
0.7% |
1.2% |
177% |
| 15 |
21 |
New Zealand |
0.7% |
0.4% |
496% |
| 16 |
17 |
Ireland |
0.6% |
0.5% |
360% |
| 17 |
30 |
South Africa |
0.6% |
0.3% |
511% |
| 18 |
16 |
Russia |
0.5% |
0.6% |
281% |
| 19 |
8 |
Spain |
0.5% |
1.3% |
127% |
| 20 |
- |
Singapore |
0.5% |
0.3% |
492% |
| Total |
89.4% |
87.6% |
315% |
What Are They Putting In Them?
| Position |
QR Code Contents |
Percentage Of Total
QR Codes (Month) |
Increase In Number
Of QR Codes Created
(Feb to Jul) |
| Jul-11 |
Feb-11 |
Jul-11 |
Feb-11 |
| 1 |
1 |
Website URL |
49.1% |
44.4% |
348% |
| 2 |
2 |
Plain Text |
25.5% |
26.0% |
309% |
| 3 |
3 |
vCard Contact Details |
11.8% |
16.3% |
229% |
| 4 |
10 |
Social Media Link |
3.5% |
2.5% |
451% |
| 5 |
4 |
Google Map Location |
1.9% |
2.4% |
245% |
| 6 |
6 |
Youtube Video Link |
1.8% |
2.4% |
244% |
| 7 |
7 |
Telephone Number |
1.5% |
1.5% |
323% |
| 8 |
8 |
SMS Message |
1.1% |
1.1% |
298% |
| 9 |
5 |
Email Message |
0.8% |
0.9% |
280% |
| 10 |
9 |
vCalendar Event |
0.7% |
0.6% |
326% |
| Total |
99.1% |
99.4% |
313% |
Methodology
I’m comfortable that these figures are representative enough to be a reasonable insight into the way things actually are:
- The numbers themselves are large enough to have a good statistical reliability – a supporting data set of 900,000 QR codes created by the 700,000 user of QRStuff.com over the 6 month period from February to July 2011.
- The QRStuff.com website has been operating since January 2008 and has had balanced historical exposure to all geographic markets (except Japan) at all stages of their take-up and maturity cycles to date. This means that the engagement timing profile for all markets is the same – been there from the start, still there now.
- QRStuff.com is not based in Europe, UK or the US, and doesn’t advertise or promote itself in any specific geographic market, so “local” bias factors would not be at play.
- My ratio of free vs paid users (87:13) would probably reflect the natural split of “DIY” and “get some informed/professional help” participants in most marketing disciplines or initiatives.
- Once I’d crunched the numbers there were no surprises or inconsistencies that were at odds with what I’d learnt, seen, heard, or suspected over the past 3 ½ years of working with QR codes commercially.
