Thursday last week, Kristoffer and I attended our final exam at BA(im) where we presented and discussed our bachelor project on Digital Patina with our supervisor and a censor.
It went very well, and we both received the highest grades possible in the Danish school system: 12.
We set out on a mission to define what Digital Patina is. Initially we had the idea that we wanted to see if the physical idea of “wearing out” could be applied to the user interface of a test website for a generic vacuum cleaner. We created three websites that were identical in content, but offered three different patina scenarios:
- Changing the size of the font from 1em to 2em
- Degrading the color from black to white
- Intensifying a shade of orange
The prograss was made in 10 steps, but we set up the site so we could mimick the effect of previous users already altering the size/color of the links, to see if our test users would pick up these signs as showing previous activity.
We figured that most of the users would recognize the effect, but the test showed us that only one user actually said “Somebody’s been here before me” and only one other user noticed any changes (but was too afraid to say it at first because she thought it was just her mind fooling around).
Basically, the test failed and applying the physical definition of patina as something being worn out doesn’t resonate with what the users expect when they enter a digital setting.
After a bit of hard thinking, we came up with what the essentials behind the wearing is: Signs of human activity. The research question reads: “How can user behaviour be shaped by digital patina?“, but instead it would have been more fitting to ask how user behaviour can be shaped by the activity of previous users, as this is essentially what we found out was the core concept behind both physical and digital interactions based on previous use.
As the paper states, the best example of an online web service shaping their users behaviour based on other users’ is Amazon’s “Customers Who Bought This Item Also Bought” as it helps me navigate towards items that I might find relevant based on what others have done.
The next best example is Last.fm’s recommendations. Because I listen to Mike Sheridan, Mikael Simpson and Melk, the web service recommends Rumpistol and is thereby shaping my behaviour.
I don’t like
Intel’s idea of Digital Patina being physical objects with RFID-tags that have some sort of data harvested through interaction with the object, as it is merely adding a digital layer to an object that already “stores” a lot of information about the people who has interacted with it. The data we know from the digital settings – in our example, a computer with a browser displaying a website – is basically dumb and objective, and we can only ever show the interaction with the data as a metaphorical representation. Our user test was conducted to see if digital patina = interface patina, but we found that the GUI is still a magical element in may people’s lives that they have learned to live with, so the patination of data needs to be more than “smudging” the interface. It works in some ways, but the user has to understand why it happens on the screen.
By giving useful metadata to the data created as a result of the interaction with it, we have reached what should define digital patina: The result of my interaction can aid future users to find what is relevant for them faster than now (or even find stuff they didn’t know they wanted until it was presented to them).
It’s long, it’s rambling, but I’ve promised a lot of people to write a bit about our paper and put it online for everybody to see, so here it is:
Digital Patina – shaping user behavior with activity patterns
Special thanks goes to Jane Mejdahl, Morten Just, Steffen T. Christensen, Anders Pollas, Kenneth Auchenberg, Aaron Bateman, Morten Kirckhoff, Britta Stougaard Mortensen and our 12 lovely testers.
