1. Introduction

The target audience of this paper is retailer chains in bikes and relating accessories.

The dataset is derived from Kaggle.com [1]. The name of dataset is “Bike sales in Europe”. The author was SADIQ SHAH, and the upload date was 2020. The dataset has 18 columns, including date, day, month, year, customer age, age group, customer gender, country, state, product category, subcategory, product, order quantity, unit cost, unit price, profit, cost, and revenue. The data ranges from 2011 to 2016. With a total of 113k customers’ information.

Bike riding has become a popular sport, both for leisure audiences and athletics. It features a healthy lifestyle, as riding can boost cardiovascular health. It is a form of aerobic exercise [2]. Also, riding is a way to improve environment [3]. With choosing riding whether for going to work or for a rest, it would not generate carbon dioxide. Goodman et al.’s [4] study found that more cycling infrastructures are helpful to promote physical activity. In recent years, with the development of modern industry, better technology has been applied on bikes, and make bikes cheaper and more comfortable to use [5]. These reasons can attract more customers to begin cycling. In brief, it contributes to a good chance to gain more profit for retailer chains in bikes and relating accessories. Moreover, Sælensminde’s study supported this opinion as the benefits of investments in cycle networks were estimated to be at least 4 to 5 times the costs [6]. Also, it is worthy to mention that men are more likely to enjoy cycling, meaning that more men can be involved in buying commodities relating to bikes [7].

Before the visualization starts, two things need to be mentioned. One is the categorization of age groups. The other is the profit margin that I calculated based on the dataset. The author of this dataset categorized age groups by four: youth (age under 25 years old), young adults (age between 25 and 34 years old), adults (age between 35 and 64 years old), seniors (age above 64 years old). For profit margin, it is calculated that dividing profit by cost.

As this paper aims at finding some methods that help retailers to gain more profits, four questions were made: What products are bought relatively more often by customers? Which customer group has the most purchasing power? Which product has the most profit margins? Will gender be a factor when designing products? These will be discussed in the following part.

2. Data visualization and analysis

Figure 1: The Heatmap showing the purchasing numbers by customers.

The order quantities of each product by customers were examined across all the timelines including in the dataset. The results are shown in figure 1, and a heatmap was used to display the relationship between order quantities and products. The darker the colour and the larger the space is, meaning that that products were bought more by customers. From this figure, the top five best sellers are concluded: tires and tubes, bottles and cages, helmets, jerseys, and caps. To generalise the common features of these best sellers, they are essential to a rider, and to highlight, tires and tubes are consumables compared to other four commodities. This means that tires and tubes need to be replaced periodically, and this can be the reason that it is preferred by riders most.

A

Figure 2: Tow bar figires showing costs and average costs by age groups

The horizontal bar figure listed the costs by age group to compare the purchasing power of different age groups. From the figure in the upper part of figure 2, it is known that adults contributed the most in bike and relating accessories sales. However, considering that the number of each age group customers may differ, the average costs by age group was compared in the lower figure. In this figure, young adults were more likely to spend more than adults in average. Also, senior customers spent not as little as the total cost showed. The number of customers from each age group did influence the total cost of them. Young adults are the age group has the most average payment costs, but the large number of adult customers contributed to the most costs.

Figure 3: The bar figure showing the profit margins by subcategories

The third step, checking which products have higher profit margins would be useful to integrate with order quantities. The figure results are shown in figure 3, and the year range is all the timeline in dataset. The top five best sellers that were generalized in figure 1 are highlighted here in this figure. It can be concluded that helmets, tires and tubes, bottles and cages feature with both high popularity and profit margins. As high profit margins with high demands can make the most profits, the final products that are most worthy to work on have been recognized.

Figure 4: The pie chart showing the percentage of customer gender in buying bike relating commodities.

Finally, the distribution of customers involved in buying bike relating commodities are shown in figure 4. The results are found similar to the phenomenon that was mentioned in the introduction. Male customers were more likely to pay more on accessories and clothing than female customers, while their costs in bikes were almost the same. These data support what Heesch et al. [7] found. It can be argued that designing products that men would like can be more profitable.

3. Conclusion

From the whole four figures, among the top five best sellers, helmets, tires and tubes, bottles and cages have relatively high profit margins. Combining the information generalized from figure 2, among the whole age groups, young adults tend to spend most on bike relating products. Adults, feature in large number of customers, and resulting in the most costs in bike relating products. Also, male customers have more purchasing power than female customers. The conclusion would be making helmets, tires and tubes, bottles and cages that attract young adults and adults can be a feasible way to make more profits. In addition, combining more elements that men would like in accessories and clothing can be a sensible choice.

There leave some problems to think about. One is that the dataset only has bike sale information (without accessory ones) in 2011 and 2012. This means that it can affect the total calculation of figure display. Also, in 2014 and 2016, only 7 months’ data were collected, this makes it impossible to check data across years by month. Another one is the definition of age groups. Over time, those who were in the prior age groups, can get older to a superior age group. This can be misleading when showing them in a multi-year figure. For the gender difference in consuming, it is also worth noting that women may not spend as much as men do, doing some design and products especially for female customers also has potential.