Myki Card Kiosk
Improving the user experience of using a Myki (Victoria, Australias travel card) Kiosk.
Tools and techniques
- Affinity Mapping
- Mental Models
- Sketching & Wireframing
- Paper Prototyping
- Usability Testing
- High Fidelity Prototype
The Myki kiosks are an integral part of the Victorian public transport system. The Myki card is used to check on and off public transport. It can be topped up online but the transaction can take up to 24 hours to go into effect. Thus a lot of people use the Myki kiosks located at train stations and many major tram stops, since topping up is instant.
The Myki kiosks do seem to suffer from a lot of user experience problems.
I set out to increase transaction speed, ease one handed use and improve the UI.
The user group of the Myki kiosks is quite broad. From my interviews and observations, most of the people using the kiosks near Flinders Street station were aged 20-40, while most of the older population seemed to have enabled the auto-top up function on the Myki website, eliminating the need to use the Myki kiosks.
I decided to make two user groups my main focus.
* The first time users of the machines, such as tourists or visitors.
* Power users who know the machines and want to do their transaction as quickly as possible.
I worked solo on this project.
I undertook all parts of the process from research to prototyping and testing.
I conducted interviews and observed people using a Myki kiosk at Flinders Street station. I also did several walkthroughs of using the machine myself.
My main findings were:
– Confirmation that the UI is confusing, with a longer than necessary transaction process
– Unclear wording
– Awkward physical layout of the machine
The context of use was also interesting to me, as I observed several people operating the machine with heavy bags on their arms, knees and back and some having trouble reaching the coin inlet.
Most also seemed to be in a hurry, suggesting the topping up the card was not planned ahead. It appeared they discovered they had to do a minute before their train was leaving.
This led me to reviewing the entire machine rather than merely sketching and wireframing
I sought to make the physical layout of the machine easier to use for people operating the machine one handed, or with heavy bags on their arms.
I also went out of my way to cut down on as many steps as possible in the top up process.
I tested my initial ideas with a full-size physical paper-based mock up. I conducted tests with the specific goal of topping up the myki card both with coins and card, and had some of my participants carry heavy bags on their arms while using the prototype.
Doing this I got a good idea of how the layout could potentially look. I wasn't confident that I had the right design yet, so to be able to rapidly test different set ups for both the physical layout of the Myki Kiosk and the UI of the screen, I created a full size paper prototype that I could conduct usability tests with. I conducted tests with the specific goal of topping up the myki card both with coins and card, and had some of my participants carry heavy bags on their arms while using the prototype.
Test feedback indicated:
– the payment method description was mistaken for a button
– my quick pay feature tested badly because it didn't fully communicate what it would actually do.
I decided to change up a few things. I connected a system change to default to a
$20 top-up when paying by card. Since this was the most common amount to top up, it seemed like an elegant solution to the problem, cutting out a button from the UI while still serving the same purpose.
High Fidelity Prototype
Having gotten good insights from testing my paper prototype, I set out to create a high fidelity design in Sketch.
Knowing that the user group was quite wide, and would also include older and not as tech-savy users, I went for a visual design that really emphasized the buttons.
As beautiful as flat design can be, it's not always good at conveying what is clickable, especially to people unfamiliar with these sorts of design patterns.
Within the limitations of the project, I think I managed to solve my initial goal quite well. The paper prototype proved especially valuable for rapid prototyping and testing. I think it helped a lot that I spent the effort to make a full sized prototype and stick it on a wall, to make the test as realistic as possible.