This work presents an empirical simulation to estimate the extent to which weather affects consumer spending. For the spending prediction purpose, the regression technique has been applied on the historical daily data of the following meteorological parameters: temperature, snow, sun, rain and humidity. The experimental results demonstrate the effectives of the model for predicting sales trends and for applying optimal business intelligence rules.
Weather has strong effects on human behavior and a lot of scientific research was devoted to the investigation on the direct link between weather and social activities. For instance, [Coh90a] and [Coh90b] reported that higher temperatures are correlated with increases in violent assaults and homicides. Also [Bar94] and [Sto99] studied the number of suicides related to the barometric pressure and demonstrated their decrease related to the wind.
Weather also influences human behavior in the sales. For instance, we buy warm clothing in winter and cool clothing in summer. Moreover, in the finance field, weather variables can affect human behavior and his mood [Goe05], [Hir03], [Sau93], [Tro97]. Coca-Cola company proposed a dynamic pricing strategy based on changes in the environment temperature, so the vending machines increase the price of a soda as the weather gets hotter [Kin00]. Nevertheless, the effect of weather on consumer spending has received only limited attention in the marketing literature [Par00] [Par01] [Ste51].
Our approach is consistent with the aforementioned studies, which analyzes the direct link between the weather variables and the customers’ behavior in order to predict the sales of a particular product in the next future. This is important to establish optimal business intelligence rules, i.e. for warehouse management. The work begins with an analysis of daily sales in one shoe store, which have an effect onweather forecasts. The paper is organized as follows: Section 2 reports an overview on the effects that weather can have on consumer behavior. Section 3 describes the technique for sales prediction based on Regression; Section 4 presents the experimental results. The conclusion of the paper and some of the most interesting future perspectives are reported in Section 5.
In literature, the weather effects have been investigated for the consumer behavior analysis. Three different consumer categories, affected by: (1) bad weather; (2) seasons and (3) his moods, were considered.
In the first category (1), the rain, the snow and extreme temperature keep people at home. In this case,the weather negatively affects both sales and store traffic[Par01].An intelligent marketing solution would be to focus on online sales. The second set of consumers (2) influences both sales volume and store traffic in particular product categories [Mau95]. For example, when temperatures fall, ice cream sales decrease, while sales of oatmeal increase. Similarity, people tend to purchase more clothing and footwear in the winter and more food and drinks in the summer [Ros00].Finally, it has been suggested that weather can influence sales by affecting consumers’ internal states (3). Although there is very little research forthis third category of effects, few studies have providedpreliminarysupport for this idea. In [Par00], the authors present a global climate-based model of the effect of weather on consumer behavior, which predicts variation in consumption patterns in response to different temperatures and exposure to sunlight. They argue that consumers do adapt to changes in the environment by modifying their purchasing behavior to both maintain physiological homeostasis and to achieve optimal stimulation levels. In [Mur10], the authors report an empirical evidence of how weather can impact consumer spending. In particular, they claimed that:
Weather variables and, sunlight in particular, affect consumer spending;
As negative affect decreases, consumer spending increases; Negative affect mediates the effect of sunlight on consumer spending.
Their research demonstrates a cause-and-effect relationship between exposure to sunlight and an increased willingness to pay for common product.
In this section, we will provide some aspects concerning the approach used for sales prediction based on weather effect.
For the purpose, we consider the empirical model for the iterative calculation of the regression coefficients: Salesij=aij+b1Tempij+b2Snowij+b3Sunij+b4Rainij+ +b5Humidij+b6SunijTempij+b7Temp2ij+eij(1) Table 1 shows the averagemeteorological values,in February 2010, related to: Temperature; Snow; Sun; Rain; Humidity.
Salesij=-13,28-1,68x1+0x2+0.44x3-0.11x4+0.17x5-0.03x6+0.16x7+1.67(2)
Table 4 describes the ANOVA parameters. An ANOVA test is a way to find out if survey or experiment results are significant. More specifically, these parameters are used to accept or to reject the null hypothesis for the alternate hypothesis. In particular: df. This measure explains degrees of freedom. SS. This measure explains the sum of squares. Regression MS. This measure is computed dividing Regression SS with Regression degrees of freedom. Residual MS. This measure is computed dividing Residual SS with Residual degrees of freedom and it explains the mean squared error. F. This measure explains overall F test for the null hypothesis. Significance F. This measure explains the significance associated P-Value.
Data of Figure 1 andFigure 4 have been calculated respect to the Equation (2). More specifically, the weather forecasts in Conversano (BA) and Milano have been used to draw the trend line, respectively.
A predicted model based on weather variables has been introduced in order to optimize business intelligence rules. The empirical simulation shows the direct link between the temperature, the snow, the sun, the rain and the humidity and the customers’ behavior in order to predict the sales of a particular fashion product in the next future. Finally, the sales prediction for two weeks has been calculated respect to the historical data in the month. In the future other models such as the “Decision Tree” in the Data Mining field will be compared with the regression approach.
The work has been developed within the framework of the project titled: Sistema di Business Intelligence di tipo “embedded” basatosu prediction real time in sistemi Big Data orientate al settore di calzature e accessorimoda “Predishoes”(Business Intelligence Embedded Systembased on real time prediction in BigData systems oriented on fashion”Predishoes”).