Inputs

Looking at the two diagrams on the left, we see that POS is the most used method by customers, with a rate of nearly 65%. Followed by POS is Payment, whose rate only equals 1/3 of of POS. The remaining methods have low usage rates, which are all below 10% and Phone Banking is used the lowest. When it comes to the distribution of merchants in Australia, NSW and VIC are the two states where most merchants locate, the proportion of merchants in there is 28% and 27% respectively. QLD ranked third with 20% of merchants and WA ranked 4th with 14%.

Looking at the two boxplots on the right, we see that both of them describe the distribution of the amount of money traded in each transaction method. Of the two plots, the upper is created from the original data and we can see that there are many outliers in this one. These outliers make it difficult for us to look at the distribution of amount for each transaction method except PAY / SALARY. This will also distort results we calculate in the following section. So we removed outliers that are way too far from the upper whisker of each boxplot. More specifically, we removed transactions of more than $5000 AUD in PAY/SALARY and over $500 AUD in the remaining methods.

Inputs

In this section, we segment dataset by date and time, then categorize it by transaction method. The two tables on the left display information of mean spending and mean volume for each month, while the two column graphs on the right display mean spending and mean volume over the course of a day. Regarding average spending, PAY / SALARY has the highest average spend for all 3 months, while the time frame 11am-17pm has the highest average spending amount. Clearly, the surge in the amount of money spent in this frame is due to employers paying employees. Of the three months, September is the month with the highest average spending, except PAY/SALARY where people pay higher in October. Regarding average volume, POS has the highest volume by month, ranging from 2500 to 2600 transactions. Followed by POS is PAYMENT, the volume of that ranges from 800 to 900 transactions. PAY / SALARY ranks third with around 300 transactions on average. Although POS has the largest number of transactions, its transaction amount is relatively low. On the other hand, PAY / SALARY has a small number of transactions which is only 1/10 of POS, but the amount of transactions is the highest. October is the highest trading month for every method, except POS where September is the highest trading month. Over the course of a day, POS is used frequently, especially at 9 am, 11 am and 9 pm. The volume of POS makes these three periods stand out. 17 pm is another distinct hour when there are higher volumes of PAY/SALARY and PAYMENT.

Inputs

These are maps of four cities in Australia in which the most stores are located. The maps on the left show the density of store locations, while the maps on the right show the density of ANZ customers’ living places. We only include customers who use POS when they make payment at stores. Looking at the density of store locations, we can see that stores in Melbourne, Sydney, Brisbane, and Perth often cluster in a certain area. For example, in Melbourne, the locations of stores are concentrated in the CBD and neighbouring high-income areas. The density of ANZ customers using POS in Melbourne is quite high, as shown by the bold colour of the contour plots, these customers live in 3 places adjacent to the CBD and St Kilda. In Sydney, ANZ POS Users focus on living near the ocean, south of the stores. In Brisbane, customers divide into 3 groups, and only one group lives near the store’s location. In Perth, customers divide into two groups who live at the edge of the shopping area.