QNT 275 Business Decision Making Project, Part 2

Abercrombie & Fitch Co. Data

Gathering for Abercrombie & Fitch for trend marketing decision making will involve the use of descriptive and inferential statistics (Solomon, Marshall, & Stuart, 2011).  Data mining will utilize existing customer sales and other information collected during the selling process, to identify statistical patterns (O’Brien, 2002).  These patterns in the data can be used to forecast trends in sales and popularity of products. The variables that need to be analyzed in the data mining method will include: customer spending and specific products. By collecting what customers are buying, the data mining tools can identify the brands and styles that are most popular.  The use of this information will provide tools for marketing decision making.

There are two primary means of research that include the descriptive statistical method and the inferential method. These two methods encompass the majority of statistical research methods. While both are useful methods they do have benefits and limitations in their use.

Descriptive Statistics

Descriptive statistics will be used to create meaningful information from quantitative data (AERD, 2014).  Descriptive statistics works best in the areas of sales and monetary data because the pure numbers are difficult to manipulate by hand and are even more difficult to visually work on (AERD, 2014). Because the data being mined is of a quantitative nature, descriptive statistics can be used to show the distribution or spread of sales data in the form of graphs.  This is especially helpful in identifying patterns in the data due to the fact that sales data such as number of sales in a specific product areas can reflect a measure of central tendency. For example, when the researchers examine the sales data they will be looking for patterns in the frequency distribution of numbers.  These patterns might include number of products sold and sales figures by days. A one month sample of sales producing scores that range from $100-$1000 with $500 being either the mode, median, or mean can be graphed reflecting what dollar volume of sales occur on which days of the month (AERD, 2014). This data shows where sales are occurring in relation to the central position. Descriptive statistics can also be used to show spread measurements such as the areas of the month that are the highest grossing for specific brands. These spreads can be measured using tools such as “range, quartiles, absolute deviation, variance and standard deviation” (AERD, 2014).

Inferential Statistics

Inferential statistics can also be used in data mining when identifying information concerning a specific population such as consumers in the under 18 market (AERD, 2014). For instance, all sales which occur with a specific brand in consumers under 18 provide an inferential statistic which isolates the marketing data for this specific population (AERD, 2014). This provides very specific decision making information for market researchers.

Trend Analysis

The use of inferential and descriptive statistics in data mining allows market researchers to utilizes existing customer sales information, spending, seasonal, and other data to identify statistical trends (AERD, 2014).  These trends, in the data, can be used to forecast sales and purchasing preferences.  These statistical tools provide more than just trend analysis for what is popular but also identifies chronological or seasonal attributes. For example, sales of winter clothing may be much higher in the winter. While this is an obvious trend, other finer or more subtle trends such as Tuesday being the largest selling day for specific clothes may indicate a trend in consumer spending.  This trend analysis can be further exploited through linear regression.

Regression is a tool used in statistics to show the dependency of a variable of one or more explanatory variables (AERD, 2014).  Regression utilizes a mathematical model that helps researchers describe or predict the relationship between dependent and independent variables (AERD, 2014).  Regression is an estimate or assumption of data based on creating a line through the data which is intended to show causality or predicting the event (AERD, 2014). For instance, by drawing a line between the dependent and independent variables of data such as money spent and brands purchased it may be possible to predict where the sales of this brand are headed. This is especially useful for forecasting sales using time series methods.

Time series regression is a forecast methods that attempts to find a linear relationship between time and the forecast variable (AERD, 2014).  Time series regression attempts to find a trend in the data that takes place over time. The time series regression requires the use of a computer and the basic formula or programming is as follows:

Y = A + BX where Y is the forecast (an unknown quantity)

A is the base or constant

B is the unit value change in the forecast variable

X is the increment of time to be forecasted

The use of time series would be especially helpful in looking for sales trends as sales occur over time. It would also be best to use a significant time period. Three years would be sufficient.

References

AERD. (2014). Descriptive and Inferential Statistics. Retrieved from AERD

Hone, F. (2007). New strategies for progressive organizations. Market forces and the new focus on demand-driven health care. Employee Benefit News, 21(1), 12.

Laudon, K. C., & Laudon, J. P. (2005). Management Information Systems: Managing the Digital Firm. NJ: Prentice Hall.

Lepore, M. (2011, April 6). Abercrombie: How a hunting and fishing store became a sex-infused teenybob legend. Retrieved from Business Insider

O’Brien, J. A. (2002). Management Information Systems: Managing Information Technology in the E-Business Enterprise, 5th ed. Boston: McGraw-Hill Irwin.

Solomon, M., Marshall, G., & Stuart, E. (2011). Marketing: Real People,Real Choices. Textbook. 7th edition. Prentice Hall: Pearson.

Leave a Reply

Your email address will not be published.