Introduction
NTUC Fair Price self-checkout kiosks are no longer a rare -sight nowadays in supermarkets. It provides an alternative payment method which reduces customers’ waiting time, while lightening the load of cashiers.
As much as these counters seem favorable, one might still notice that the elderly tend to avoid it and the younger group sometimes grumble about its user friendliness.
Our objective is to uncover these issues and discover insights through Design Thinking, a five-stage creative problem solving process, in order to make improvements to the UX design of these kiosks.
We started with the Empathize process where we tried to understand how users (us included) think and feel about the self-checkout kiosk. We used the existing self-checkout kiosk in Fair Price to understand the current process and document our observations and thoughts on the pros and cons of the user experience.
Immerse: We started off by immersing ourselves as the role of a user, and conducted shadowing, contextual inquiry and interviewing of other users. We put ourselves in the position of users and documented our thoughts and feelings as we navigated the current self-checkout kiosk. We consolidated our findings on each user’s empathy map below.
Observe: As we analysed the data collected, these are some pain points that surfaced up more prominently: (1) Lack of language choices, (2) users are not able to go one step back, (3) users are not able find the options they need from the many options provided.
These information are then consolidated into the Aggregated Empathy Map below.
Our target demographic for this project is the 30–50 years old age group, as we have observed them to be the most active group that does weekly grocery shopping. In addition, this group is more likely to use self-checkout kiosks.
From the data gathered on the Aggregated Empathy Map, we started to consolidate and categorise them into the Affinity Map (above) to discover more insights.
Our key insights are split into 4 categories:
- Requires a lot of staff assistance
- A lot of confusion with the instructions given
- Insufficient space for packing groceries
- The stress of holding the kiosk for too long
With these key insights, we sorted them into 4 sections below. This gave a clearer objective in creating our persona based on our demographic’s needs and frustrations.
Meet Cecilia, a working mother who loves to get the best deals “lobang” for her groceries and highly values efficiency.
Next step is to understand our persona’s pain points on her checkout journey. Our customer journey map (below) allows us to discover Cecilia’s frustrations while making payment at NTUC Fair Price on a typical day.
Summary of what we have observed that caused dissastifaction:
- Having to queue for a long time to use the self-checkout kiosk even though she only has a few items in her basket.
- Having to ask for assistance due to confusing instructions on screen but there is only 1 staff in charged of 6 kiosks thus it means Cecilia has to wait again.
- Staff didn’t explain clearly after solving the issue which Cecilia encountered. This made Cecilia irritated because she wanted to learn so she doesn’t make the same mistakes again.
Brainstorming, a crucial part of design thinking, takes place in this ideate section. The team was mentally prepared to “go wild” in terms of concepts and outcomes. We have a mindset of “It’s not about coming up with the right idea, its about generating a range of possibilities”.
Method 1: HMW (How Might We)
Based on our persona’s frustrations, we came up with a bunch of HMW statements to kickstart a wide range of possible solutions addressing that.
Method 2: Lotus Blossom
We expanded one issue: “How might we make the whole checkout experience more intuitive” and expanded our ideas using Lotus Blossom and Scamper. We decided to address this point because of the pain points identified from the customer journey map.
Still keeping our persona’s pain points in mind, we decided to focus on creating a seamless experience at the self-checkout kiosk through these 3 aspects:
- Add graphics to instructions to make them clearer,
- More user-friendly checkout flow,
- Speed up time taken to scan items and payment.
Method 3: Scamper
The team further expanded the solutions from ‘Lotus Blossom’ to ‘Scamper’ method (below). This technique helps us to generate more ideas and possibilities. We continued to brainstorm for more ideas addressing our persona’s pain points.
Method 4: Impossible to Possible
This method frees our mind further from all the logical possibilities we have previously considered. It was fun to come out with crazily stupid impossible ideas before refining it into the possible ideas. We then realised our ideas at the “Possible” stage were not thought of in other brainstorming methods above.
Method 5: Relevance and Novelty
Lastly, with so much information thought out, we picked a handful of ideas and categorised them into these 2 categories (below); Relevance and Novelty.
These ideas will then be further shortlisted by the team (one from Relevance side and one from Novelty side) to be included as features in our prototype.
By understanding our target audience based on our findings in empathising with the users, defining the problem and ideating new ways to improve the flow of the checkout process, we next moved to the prototype stage. This stage allows us to test our ideas quickly and improve them by using prototypes.
This is where the fun begins. The team drew the first draft to test a number of ideas without investing so much time upfront. After several discussions, we produced a second round of drawings to showcase the flow below.
This is our first low-fidelity (non interactive) prototype. As there are many loop holes and uprising problems in terms of the flow of the process, many failures, and many drafts were made.
Our first site map above shows the key items the team has decided to implement on our new checkout process which includes at least 1 idea from ‘Novelty’ and ‘Relevance’ respectively. Here are the features the team has selected for this prototype:
- The power of using face recognition for sign-in as member.
- There will be 4 languages to choose from.
- Animated videos instructing users on what to do.
- Automated self scanning one basket at a go.
- Adding a QR code payment.
We re-drew all our drafts in detail on an iPad using Procreate application. Due to the current pandemic, we decided to make our prototype interactive by using the Marvel application.
The planning process on the flow of the Mid-fidelity (interactive) prototype included many rounds of trial and error, layout and placement considerations. Refinements were proposed within the team and applied to our prototype. Every step has to be well thought of by putting ourselves into the user’s shoes and understanding how the users think and navigate through the process.
Now everything comes together and that is when we will know whether our research, observations, ideas works. Our team came up with 3 scenarios and tested with 5 users that fall within our targeted audience. Below are three scenarios created for user testing:
- Scenario 1 (“Indiana” John)
NTUC self-checkout kiosks introduced a new face recognition function for members. Being curious and adventurous, you decided to give it a try and make payment with EZ link card since you haven’t done it before.
User’s Task: Sign in as a member using the face recognition function and make a purchase using EZ link card. - Scenario 2 (“Sotong” Jane)
You are a NTUC member but you have forgotten to bring your membership card. Thus, you decided to use your mobile number to sign in and you are limited to make your payment via NETs. Midway through the process of checking out, you realized that you don’t need an item from the list and decided to remove it. You decide to email the receipt to yourself.
User’s Task: Sign in as a member using your mobile number and make a purchase using NETs. Try the function to remove Nestle Kit Kat from your list of items. Email receipt to yourself. - Scenario 3 (Husband Jim)
You are tasked to do a last minute errand by your wife and you have to get them quickly at the nearby supermarket. You managed to find what she wants and made the purchase using your credit card under as a non-member. As a proof that you didn’t loiter around, you emailed the receipt to your wife as a form of evidence.
User’s Task: Sign in as a non-member and make a purchase using credit card. Email receipt to wife.
The Good, The Bad, and The Possibilities
After all 5 users are done testing the prototype, we asked for their feedbacks on areas that are good, need improvements or any potential areas that the team might have overlooked.
Conclusion
Solutions proposed in regards to the findings and insights were well received by the users at user testing stage. 10 users in total including all team members were able to complete each task swiftly with minimum task flow difficulties.
Overall, the users find the app an improvement from the current, especially in terms of speed of transaction and user-friendliness. However, they still find that some buttons can be placed in a more prominent position to be noticed and some security issues in regard to membership sign-ins. Users also find the Novelty face recognition function to be a good concept and sees its potential.
Design thinking doesn’t end here, it repeats in order to make the product better. But, we have to end our report here. Thanks for reading.