Darin Materman, Bea Tugano, Michael Mehall
Understanding how purchase behavior is influenced by various factors such as location, promotions, demographics, and loyalty is crucial for businesses aiming to optimize their marketing and sales strategies. Our analysis seeks to examine how these elements impact both the frequency of purchases and the size of baskets, providing insights into consumer decision-making processes. By examining these factors, we can glean insights into how businesses can tailor their approaches to better meet the needs and preferences of their target audience, in order to drive sales and increase customer satisfaction and loyalty. Succinctly, the question we’re looking to examine is: How is purchase behavior (frequency and basket size) impacted by buyer location, promotions, demographics, and loyalty?
The dataset is from Kaggle originally composed by Sir Sourav Banerjee at CogniTensor and last updated a year ago. It contains two tables covering demographic information, purchase history, product preference, and preferred shopping channels with the first table focusing on customer behavior while the second on customer trends. Both contain about 3900 observations across 18 columns, including both categorical (such as gender, size, season) and continuous variables (purchase amount, age). Data is licensed under Creative Commons and thus is free to use.
Our dataset on Retail and Consumer Purchase Behavior can be found here .
Description of the figure: This figure is a bar chart of the most preferred categories by shoppers. The figure shows the distribution of most commonly purchased categories across the whole dataset. Clothing is the most popular category while outerwear is the least popular category.
Description of the figure: This figure is a histogram of the count of purchasers by age. Each bin represents five years, and the elevent buckets encompass the whole dataset. Naturally, there are less younger buyer, however the count remains remarkably stable between ages 25 and 70. A scatter plot did not reveal any relationship between age and purchase amount. That being said, our plot is valuable because it highlights the applicability of the dataset being that we capture such a wide age range.
Description of the figure: This figure is a histogram of the count of basket size across the dataset. Ranging from $20 to $100, the figure indicates a small parabolic trend, where more purchases occur at lower basket sizes ($20-$40), and higher basket sizes ($85+), while baskets in the middle occur with less frequency. This informs us that customers are often making large or small purchases, putting one or two things in their cart, or many.
Description of the figure: This chart shows the items purchased by season, where each item is colored by category. The viewer is able to switch between seasons or view them in the aggregate in order to drill down onto more specific data. This stacked bar chart shows that generally, clothing are the highest purchased items, making up four out of the five most purchased items across all seasons. As one would expect, the outerwear category sees higher sales in the Fall and Winter compared to the summer and fall. That being said, the dataset also comes with some key surprises. For instance, Pants are the leading item in the summer, and skirt is one of the lowest selling items of the summer. This is potentially due to people planning their shopping ahead (buying pants for the fall because they already bought skirts in the spring, which is corroborated in the data.)
Description of the figure: This boxplot visualizes the distribution of item purchases by season, with each item categorized by color. The chart reveals that, overall, accessories tend to have the highest purchase counts, particularly backpacks and belts. Hovering over each box reveals distribution information for that item, and hovering over each outlier tells you which season it comes from. Notably, the outerwear category sees a spike in sales during the Fall and Winter seasons, as expected. However, there are some surprising insights, such as the fact that T-shirts sell poorly in the summer, while sweaters are consistently popular throughout the year, especially in Winter. These patterns may suggest that consumers are purchasing items in anticipation of the coming seasons and helps build context to the bar chart above.
Description of the figure: The chart on the left shows the average spending across all categories by gender. It can be seen that females spend on average more than males by about $1. However, when looking at the broken up bar chart on the right, the categories that females and males primarily spend on are different. While females spend the most on accessories, followed by clothing, footwear, and then finally outerwear, males spend the most in footwear, followed by clothing, accessories, and finally footwear. These differences can be observed by selecting either a section or bar of the charts to highlight just that category, or by selecting a gender by clicking on the grey square in the center legend. This allows for an easier comparison between categories in the same gender, or the same category.
The conclusions of this chart illistrate that stores should focus on certain categories for each gender. For females accessories should be prioritized while males spend more on footwear. Outerwear is the least spent category for both, thus indicating that it should be minimized in comparison to the other categories.
Description of the figure: This figure shows two side by side choropleth maps of the United States with states colored on the average spend and purcahse frequency. Hovering over each state reveals more information about the average spend as well as the frequecy of shopping. States like Colorado and Wisconsin have both high frequency and low average basket price, meaning those are uniquely good states to target to increase average basket size. Alaska, for instance, has high average spending, but below average purchase frequency, meaning they are a good target for increasing basket size. This map shows places like Utah, Washington, and North Dakota as the ideal states, with above average purchase frequecy and basket spend.
This analysis of consumer buying behavior has provided valuable insights into the impact of key drivers such as location, promotions, demographics, and loyalty on both the frequency of purchases and basket size. By examining the dataset, we were able to identify several significant trends that can inform how businesses tailor their marketing strategies and optimize their sales processes.
Impact of Demographics on Consumer Behavior
The relationship between consumer demographics (such as gender and age) and purchase behavior was strikingly evident. Our analysis showed that while males tend to spend more overall, their spending patterns across categories are consistent with those of females, with clothing being the dominant category. This suggests that while demographic differences are important, the broad categories of interest remain similar across genders. Furthermore, we observed that younger consumers were underrepresented in the dataset, which could suggest that marketing efforts should be adjusted to better target this group through digital channels or tailored product offerings.
Role of Promotions in Shaping Purchase Behavior
Promotions played a significant role in influencing both the frequency of purchases and the size of baskets. The data demonstrated a parabolic trend, where a higher frequency of purchases occurred in lower and higher basket sizes, with fewer purchases made in the mid-range. This suggests that promotions can be highly effective in driving either low-cost, frequent impulse purchases or larger, more considered transactions. For retailers, understanding these patterns can help in designing promotions that align with consumer behavior, such as offering discounts for bulk purchases or strategically timed sales that encourage larger baskets.
Seasonal Trends and Product Preferences
Our exploration of seasonal trends revealed interesting dynamics in how product categories are purchased throughout the year. As expected, clothing remained the most consistently purchased item, with outerwear peaking in the colder months. However, the summer months saw an unexpected rise in pants sales, indicating that consumers may be purchasing for future seasons. This highlights the importance of anticipating consumer needs and adjusting inventory and marketing strategies accordingly. For businesses, understanding these trends allows them to better predict demand and optimize stock levels, reducing overstocking or understocking issues.
Consumer Loyalty and Long-Term Engagement
Loyalty programs and their long-term effects on consumer behavior remain critical components of a business’s strategy. While our data did not provide a direct measure of loyalty program participation, the trends observed suggest that loyal customers are likely to spend more over time, particularly in categories they prefer. This aligns with existing literature on the importance of loyalty in fostering repeat business. For companies, designing effective loyalty programs that reward frequent shoppers can lead to an increase in both the frequency and size of purchases.
Key Insights for Businesses
Our analysis offers several practical recommendations for businesses looking to optimize their strategies:
Future Questions and Analysis
While this analysis offers significant insights into the factors influencing consumer buying behavior, there are several areas where further research could be beneficial:
In conclusion, while this analysis provides actionable insights for businesses to refine their marketing strategies, the evolving nature of consumer behavior and market conditions suggests that continued research is essential. Future studies will help businesses stay ahead of trends, improve customer experiences, and ultimately drive higher sales and long-term loyalty.
Frontiers | Factors Affecting Impulse Buying Behavior of
Consumers
This study examines various factors that influence impulse buying
behavior, which can be related to purchase frequency and basket
size.
Customer Purchase Behavior Prediction in E-commerce: A
Conceptual Framework and Research Agenda | SpringerLink
This paper provides a framework for predicting customer purchase
behavior in e-commerce, including the impact of location,
promotions, demographics, and loyalty.
The Long-Term Impact of Loyalty Programs on Consumer Purchase
Behavior and Loyalty
This study examines the long-term effects of loyalty programs on
consumer behavior, including how these programs influence purchase
frequency and basket size over time.