This analysis looked at three features of a Kaggle dataset that included the most recent Craigslist used vehicle listings to help used car dealers make better business decisions. First, data preprocessing was done to make sure there were no invalid data. All data processing work was done in Python, and several visualizations were utilized to demonstrate the findings.
The data preprocessing step reveals some basic market information, such as the typical year of manufacture, odometer reading, and average price. This is valuable information for dealers since this information enables dealers to identify their potential customers and offer precedence to those customers who are searching for these kinds of vehicles. Dealers are more likely to complete the transaction quickly if they can identify the target clients.
The preliminary analysis concentrates on the national scale. The top 5 most popular brands, the most typical year of manufacture for used cars, and the most prevalent vehicle types were examined. According to this section, Toyota, Honda, Nissan, Ford, and Chevrolet are the most popular and widely available brands on a national scale. The most popular year of manufacture for used vehicles on the market is 2013. In terms of used vehicle body types, sedans and SUVs dominate. The dealers, especially those who operate businesses throughout the nation, can have a decent sense of the kind of used car they should acquire according to this preliminary analysis. Acquiring those types of used cars is profitable since they are in high demand and readily accessible in the market, so dealers are more likely to buy them for a reasonably low price and sell them quickly. It is also important to note that, even though this preliminary study does not demonstrate the rising popularity of electric vehicles, given that electric cars are still relatively new to the market, dealers should think about acquiring them to boost sales. This is because more and more people are choosing electric vehicles due to their technology and affordability.
The following analysis focuses on the regional level. This section investigates the average price distribution of used cars across the country and the preference of brands across the country. This section reveals that average used car prices tend to be higher in major cities on the east/west coast and the Midwest. However, average used car prices decrease significantly in less populous areas, even though those areas are only a hundred miles away from metropolitan areas. The preference of make was also investigated and the result suggests that Toyota is the most popular brand in the used automobile market in most urban areas, while Ford or Chevrolet are preferred in less populous areas. As a result, vehicles such as Toyotas will be reasonably priced in remote areas. This means that dealers may benefit from this chance for arbitrage by buying Toyotas in rural areas of the same state at a lesser cost and then selling them in urban areas at a higher cost, and vice versa. Since shipment takes place within the state, the profit is very much ensured.
The relationships between numerous subtler features and prices were analyzed by plotting two heatmaps, namely the heatmap of price contributors and the heatmap of colors, to specify further the precise type of vehicles the dealers should prioritize when making the purchase for the aforementioned arbitrage. The first section reveals that a like-new, 4-wheel drive, full-size car with low odometer reading can positively affect its price. The second part shows that prices are not much affected by the car's color. White is the only paint color having a comparatively substantial positive relationship with pricing. For dealers in metropolitan regions, the most lucrative choice to buy will be a white, like-new, 4WD Toyota car with low mileage listed in the rural area. Additionally, dealers in rural locations might make the greatest money if they buy a white, nearly-new Ford vehicle with low miles that is being sold in an urban area.
As a result, the dealer will have a greater awareness of the market and consumer preferences thanks to this project. In this report, interactive visualizations were extensively used to enhance understanding of insights, so even dealers who have never seen this type of research before may quickly comprehend the main idea and utilize the tactics in their daily business decision-making process. In addition to visualizations, concrete examples were provided in each section to assist the dealer in better understanding the business concept. Profits are therefore guaranteed by their capacity to implement those tactics daily. In future work, one may utilize the established plots for feature selection and build a sophisticated pricing model to help the dealers quantify the profit gains.