top of page

Current Issue 

The problem centers around how customers prefer to use the service from the in-store staff despite the installment of self-service kiosks. 

Project Goals

The main goal of this project is to simplify, streamline and reimagine the self-service user experience for the customers to opt for the kiosk instead of relying on the service counter staff. Hence, the team's execution will mainly focus on the kiosk experience. 

Competitive Analysis

To find the best approach in solving the self-service issues, my team and I decided to do some initial desk research about some potential competitors to gain insights of the product in the market. We decided to look into USPS, McDonald's, airlines, and some supermarket kiosk services. 

Defining the issues

To further understand user wants and needs of the self-service kiosks, my team and I decided to conduct on-site research by visiting Australia Post office and observed the current in-store kiosks. By doing so, we can identify the core issues with the current sending experience and the in-store friction between the users and kiosks. 

We spent about an hour shadowing the activity at the site and recorded the results.

  •  More than 20 people were willing to use the kiosk when there was no long queue at the service counter. However, only one did not ask the staff for help, while the rest needed an assistant. 

  • The distance between the parcel packing area and kiosk is pretty far; users have to travel back and forth between stations several times. 

  • The majority needed help from the staff when using the kiosks and got stuck immediately on the first screen.

  •  A pattern of self-scanning the parcels by themselves and immediately asking for help 

We also tried to use the kiosk by ourselves; below are the factors we listed out that makes the experience unpleasant for the users. 

  • Unclear instructions

  • Unorganized layout that isn't user friendly, crowded screens

  • Excessive and also irrelevant text that overwhelm the users

  • No system- send parcel and send letter services are merged together into one

  • No hierarchy in the user interface

Though if one manages to complete the sending experience on the kiosk, they will feel uncertain as there is no confirmation and a lack of transparency of the process from the machine.


We even spotted a measurement tool, a letter gauge slot, attached beside the kiosk. However, because it is not indicated, users often ignore them. 

Interviews and Surveys

After completing the initial observation of the kiosk, we also conducted several interviews and surveys. We interviewed people who used Australia Post sending service and obtained valuable customer insights such as their opinions and pain points. 

Thirty participants undertake our online surveys. Moreover, to pinpoint the issue as accurately as possible, we categorized the survey questions into groups ( as seen below) to make sure we tackle all of the topics we planned to get deeper insights.

Main Findings

  • With 80% claiming that they are tech-savvy, 70% would not choose to use the kiosk and instead ask staff for help. 

  •     The most frustrating part at the post office is the long queues of line and long waiting time. 

  •     Their ideal sending experience is a simple process and track options availability. 

  •     55% said their kiosk usage is for sending parcel and 15% is for sending letter. While some use it to pay bills and pay stamps. 

  •     50/50 split result with pre-packaged or do in-store packaging

  •     Distance between the packaging area and kiosk does not cause user dissatisfaction, but it doesn't hurt to be improved.

  •     65% claimed they never wait at the kiosk as no one opts for the kiosk service. 

  •     The best part of the kiosk experience is the absence of queuing; additionally, the worst part of the kiosk is its complicated user interface which hinders them from completing the task by themselves.

  •     Some even claimed the kiosk's interface lagged a lot; the worst-case scenario is that the machine doesn't work. 

To conclude, the purpose of the kiosk is to introduce automation that eases the sending experience. Yet, with the responses obtained, it can be said that the automation service is currently a hindrance. 

Mid Fidelity Usability Test

We then conducted a mid-fidelity usability test for feedback and further refinement. The goal was to 

  1. Successfully enables customers to guide themselves to complete a task independently

  2. Ease their in-store experience with the newly designed kiosk interface

We categorized user testing into two different tasks and three different scenarios for 29 participants. It allows us to see whether we have achieved in making the customers understand various features implemented onto the kiosk. Moreover, participants age ranging from 18-60 are asked to participate.

As for the degree of the testing, we decided to conduct a Retrospective Think Aloud(RTA) and Restrospective Probing( RP) at the end. One of our goals is to reduce the time it takes for one to complete a task. RTA technique gave us accurate time on task duration and the user's natural behavior using the product. Moreover, by asking the participants to retrace their steps, we can evaluate whether the experience is memorable for the users to recall their decisions. 


As for the qualitative data, we ask questions about their actions during the task to find room for improvement based on user insights that could not be collected by performing ourselves( as biased might be involved). 


Test the mid-fidelity prototype here

Maze was used to conduct our remote testing to gain quantitative and even qualitative data from our extensive samples, as the platform can break down all the testing into numerical values and percentages. 

We can see the number of users who followed the direct and indirect paths and those who gave up on the screen. These results help us identify the screens that reduce user satisfaction and not make any risk-tasking design iterations. 

The platform also automatically calculates each participant's score when they completed the test and the misclick rate performed on each screen. Last but not least, the average time participants spent on each screen are also recorded. 

Usability Breakdown 

After seeing the testing results on Maze, we were surprised by the various outcome we received. While some manage to complete the task fast, some struggle with the section; others find it easy. The usability score on each individual differs dramatically, fortunately, we set a few questions at the end for users to answer which most of the participants find the kiosk easy to follow.

We later realized that when we set the user flow for a direct path on Maze, we did not selected an experience that users usually click. ( for example, we didn’t make the toggle interactable, so when the user clicked them, it add- ed the misclick rate.

Still, after reading their answer and seeing the results sum- mary on Maze, we can say, for different scenarios, users successfully completed their assigned tasks despite the confusion in some sections (i.e., address). With that said, there are still rooms for refinements . 

Major Issues

  • Additional interface features distract/ interrupt the user to understand the primary step on the screen (lack of interactivity for the keyboard function) 

  • There are still choppy jumps between screens; users expect something more detailed and seamless. (i.e. address page) 

  • Some text on the buttons is too vague, which confused the user.

  • Buttons are placed in a location that causes users to misclick them constantly.

  • Some screens do not have a clear hierarchy. 

High Fidelity 

We rewrote several UX writing during our final prototype stage based on the feedback we received during the user testing. Instructions and text buttons are written with a clear message for users to understand the visual feedback after interacting. Screens are designed with no jumpy user experience. Elements on each screen are adjusted to avoid off-track user flow and present a clear visual hierarchy for user satisfaction. We designed the entire kiosk flow with new and returning customers in our minds by creating the screens as easy to follow for them to complete their tasks independently. 

Additional Adjustments

  • Handwriting features for address info input

  • Progression circle charts are added for customers to know how many steps are left to complete, transparency of the service.

  • Icons and graphics are added for easier understanding.

  • Add language setting 

  • Additional screens that calculate real-time weight for the package

  • Keyboards are hidden when not in use.

  • Pop-up info appear when the information tab is clicked for users to understand the context better and avoid selecting the wrong options.

  • Visual Instruction of the next step after completing the payment is added: peel the label, stick on the package, and drop it off. 

  • Cancel buttons are accessible on every screen, which return the users to the home screen

User Testing

​The high fidelity testing was conducted in a moderate face-to-face manner where participants completed a task through Invision. There were 7 participants in this test, and the result showed that all of them easily completed their assigned tasks. We also ask them to recap some aspects they like about the product. There are positive reactions to the addition of a progress circle bar, handwriting address feature, and confirmation pop-up. 

As a team, we were surprised by the participants' responses to the handwriting features. We think it is a new feature that the customers need time to be familiarised with; however, we received feedback that it saves their time and is a fun addition to the interface. They also praised the color usage of the product cards, where we use different colors to identify other product options. Overall, the high-fidelity prototype exceeds the performance we think could perform for the users. 

High Fidelity Usability Report



Final kiosk deliverables

The self-serve kiosk could be separated into two user flows: 

  • sending a parcel

  • sending a letter

As the flow of these two sending experiences is different, we decided to separate them on the home page. For a parcel sent out, customers must place the package on the scale, whereas for the letters, they must measure the dimension on the gauge provided before weighting their letters.


Additional proposal

As our intention is not solely to improve the self-serve kiosk service but also the in-store experience, we present a draft of how the kiosk section should be arranged for convenient usage and to enhance agility. In addition, we proposed how the letter gauge should be labeled to prevent customer misunderstanding. Hence, the measurement tool that was initially implemented will not be overlooked and will be used for its purpose. 


We got a chance to propose our ideas to Australia Post and were selected as the best execution for the project. From the feedback received, the factor that our team got chosen is related to the highly attainable concept that we had when compared to other teams. We brought what was already there (such as the letter gauge) to the customer and did not take the overhaul approach of creating a product that is completely new.

Moreover, the additional suggestion of rearranging the kiosk setup was intriguing as minor adjustments can significantly change the in-store experience. In conclusion, the whole designing journey that my team and I took on Australia Post had so much potential and was selected as the ideas that would be further developed. 

Self Service Kiosk

Project Summary :

Reimagining customer in-store experience at Australia Post with redesigning self-service kiosks system interface and process as a primary focus

My Role 

User Research, Usability Testing, and prototyping the deliverables 


In-store Kiosk- Australian Post Office


Adobe XD,  Invision,  Maze,  Adobe Illustrator



Paper Sketch

We brainstorm the elements, items, and flows on the kiosk system interface. Listing out the main screens that must be included in the kiosk system, that way for further iterations, we will have a clear baseline to elaborate upon the main screens. 


What I learn from the experience

The outcome outperformed what we imagined and received lots of positive feedback from Australia Post. However, if I had a chance to improve the project, it would be on the usability testing. We could gain valuable insights from the users if the test was conducted face-to-face. I would observe the interaction of the users while they were testing the prototypes and record any unexpected or indirect decisions they made and ask them at the end. As a result, we could better understand their thinking process behind those intuitive actions and further elevate the kiosk user experience. 

Additional User Testing

Initially, our goal as a team was to simplify the kiosk interface; however, as the project process, we realized complexity is crucial to delivering clarity. We can streamline the visual appearance and the flow to just a certain extent as there are many essential screens and options that we are dealing with on the kiosks. From this experience, I learn that simplifying outcomes does not always result in a positive solution as clarity is also taken away. 

Beware of oversimplifying

Top Graduates Platform
bottom of page