Mobile QR codes User behaviour

Early App and QR code scanning data from Love Lace exhibition

I promised updates on the data coming from the QR code implementation in the Love Lace exhibition so here are the results of the last 4 weeks since opening.

Already we’ve released updates to both the iOS and Android versions of the Love Lace App. Perhaps surprisingly it has been the Android App that has given us the most trouble. Carlos has been troubleshooting various Android devices and OS versions to make the QR code scanning work properly – something that has been made much easier on iOS because of the consistency of hardware and lockdown of other apps. Now, though both are humming along nicely.

In terms of downloads we’ve had 572 iOS and 165 Androids. And using Flurry we’ve tracked 3,126 sessions on iOS and 502 on Android.

But let’s jump to the meaty data.

When we designed this App the QR code scanning tool was built in to try to maximise the use of QR code scans in the exhibition. Of course users could still just browse the scrolling list of objects and artists if they wished, but we hoped to get the QR scanning up to a reasonably good level by reducing user friction.

Looking only at the iOS figures we can see that browsing is by far the preferred behaviour although we haven’t segmented this by location. Obviously the QR code scanning only works when the visitor is in the gallery and outside of the gallery any App use would involve the scrolling browser only.

233 items (objects and artist records) have been viewed a total of 6933 times using the scrolling interface.

The QR code scanner has had 844 scans including 45 failed scans and 17 non-exhibition codes. Many objects have not been scanned at all.

Where this becomes interesting at this early stage is when we overlay the scans on the exhibition floor plan.

(click to open this at full size in a new window/tab – 457kb)

Visitors enter this gallery space from the bottom left and then complete a circuit counter clockwise. The triangular grey area in the very bottom left is the exhibition title wall that has signed promoting the App and the free in-gallery wifi.

Not unexpectedly the first hemisphere of Room 1 followed by Room 2 attract the most scans. However after that things become interesting.

What is striking about the overlay is that the most popular object (Meghan Price’s Habitat Wave) is near the end of the circuit of this part of the gallery in Room 8 and this is a rare outlier, being surrounded by almost entirely unscanned objects. Similarly Room 6, full of smaller objects, has a cluster of scanned objects but these are comparatively low numbers.

The cluster at the top of Room 10 are a set of five QR codes linking to the Inter Lace microdoumentaries that are projected in a remixed form in this space. Visitors dwell for significant time in this area but from the low figures would not seem to be aware of the full versions of these documentaries that lie in wait on YouTube.


In the next few weeks we will be rolling out a newer version of the App which will incorporate both these documentary videos as well as the ability to ‘love’ objects and share them more easily. We will be able to compare this data with the scan and view data and see if there are any correlations. Then, in about six weeks time the moveME wifi triangulation system will also be integrated allowing us to overlay and correlate dwell times in the space against ‘actions’ such as ‘love’ or scanning.

Stay tuned for the results of that.

11 replies on “Early App and QR code scanning data from Love Lace exhibition”

Seb, this is great information.  Thanks!  how many visitors has the exhibit had – that is, what percentage of your visitors have used this technology? 

We don’t have an entry count for this particular gallery just yet although that will come along with the wifi triangulation in the next install. From my observations that take up of the App and QR codes has been very low – maybe under 1%. 

Although we  are also looking at ways of addressing this by putting the ‘How To’ App video on an iPad or screen prominently in the entrance way.

Thanks – we’re a game museum and just came back from a major games expo – I suspect our demographic would be more receptive to these technologies, so I’ll be following this closely in the next year with the hope of implementing it at the same expo next year.  I hope we can get some exhibits up at local venues – we’re in Silicon Valley, and if people don’t respond to the technology here, I’ll be very surprised.

We really wanted to see how this developed over time as people came to the exhibition and used it. We outsourced or scaled back a number of parallel projects to ensure that this one, which offers significant internal learnings was able to flourish. That and Carlos, our developer’s desire to learn something new and test some code ideas. We also had the previous QR code work to draw upon and improve.

We have implemented QR codes in our Israeli Innovation exhibition at the Bloomfield Science Museum Jerusalem. Near the logo of each company display, we have published a QR code that leads to the website of the company.

This was a first for us.

I am now completing the report on this experiment.

Inspired by this very post, I have also created a  “geographic analysis” chart that gave me similar results.

I think that we can coin a new term – “QR Fatigue”.

I really appreciate the statistics and the detailed account of this “experiment” — I have been gathering data of my own using my app, Mused for iPhone (, which has a QR scanning feature (in the unreleased 1.5 version currently in very limited distribution) specifically used for scanning museum QR codes (though it will technically scan any).

For museums, my goal is to make it easy to print QR codes, see statistics on number of scans, durations between scans, and other interesting data points. Every museum will have a different deployment strategy, and thus will have very different data and user activity. It will be interesting to see how it differs.

For consumers, my goal is to make it easy to scan QR codes at any museum, and not necessarily require a custom application for each museum (though not exclude users from it… perhaps finding ways to link in to their applications and giving the users a choice to use it). This will be especially good because smaller museums will not want to build their own applications but will still want to take advantage of the benefits QR codes have for their exhibits.

Really it’s still just an experiment for me too. Consumers and museums will be the biggest drivers for how QR codes will be harnessed moving forward, and I’m excited to see what happens!

I am very anxious to see how the triangulation works for you.  I have been looking at near field communication (NFC) which is to be embedded into mobile devices beginning in 2012.  I am wondering if you are also looking at this technique for location-based content delivery or visitor tracking?  I would be curious to know the pros and cons of each method.

Hi Linda

I’ve looked at NFC before and back at Picnic 2008 in Amsterdam it was all the rage. Nokia had their new phones out then with NFC but then the iPhone happened and it didn’t have NFC. Whilst NFC is almost certainly going to come to more smartphones in the next two years, I expect there will be at least a few years – assuming it goes smoothly – before it gets anywhere near enough consumer take-up and familiarity to be useful.

And by that time I’d expect that visual recognition – Google Goggles, ArtFinder etc – will all have made significant progress towards improving their recognition of 3D objects through the camera. Given that 2D works pretty well now, I’d wager that visual recognition is probably the ways things will head.

So, if I was a museum investing heavily I’d be looking at lightweight solutions now, and focus all the resource and effort at getting the content and data ready so that whatever technology ends up getting mass market acceptance has something to deliver.

That’s why WiFi seems like a safe bet now – as you are going to have to do that in any case. And if triangulation gives a ‘good enough’ outcome in the short term at reasonable cost, then you can experiment with content and getting your big datasets ready.

Leave a Reply

Your email address will not be published. Required fields are marked *