Every day, e-businesses collect huge amounts of data from their e-commerce Web sites. The raw data alone has little value, so how can an e-business effectively leverage the data from its site? It can identify and collect the information that management needs, consolidate it in a data warehouse, and then analyze it for leveraging opportunities.
The E-commerce Data Pool E-commerce Web sites can collect a wide variety of data:
• Site traffic metrics, including how many visitors view the site during a given time period, what parts of the site they visit, and whether visitors return
• Sales statistics for the various products sold on the site
• Customer profile details, including age and geographic distribution
• Customer buying habits
Like any data collection effort, quantity doesn’t mean quality. The first step is to ensure the site collects the raw data that management wants to analyze. After identifying those requirements, the Web site’s business logic can be designed so the required data is edited and collected each time a customer completes a transaction. It is critical that the Web site contain the necessary business logic so that the raw data for the warehouse is properly collected.
Analyzing the Collected Data
After collecting the required data, management can establish criteria, analytical approaches, and systems to manage, analyze, and leverage it. For example, a simple analysis may show the majority of customers who buy a particular product are male and are in the 20- to 30- year age group. With this information, management can target its efforts toward that particular customer segment and increase product sales.
Various techniques exist for analyzing the data collected in the warehouse. Often, user requirements dictate the proper analysis technique. SQL can be used to examine the data, and raw statistical information can be compiled. This classic technique is best for the study of detailed customer transactions.
Another approach is multidimensional analysis, commonly referred to as online analytical processing (OLAP). Using multidimensional analysis, users can perform their own custom analyses and view data from various perspectives. Multidimensional analysis tools permit users to drill down to a detailed data level from a more summarized level and vice versa. There are a number of OLAP products for multi-dimensional analysis.
Another technique is data mining. Data mining algorithms can be applied to find patterns hidden in the data. The problem with applying data-mining algorithms is if the data is not accurate, the results will not be accurate.
Simplifying and Improving the Analysis
A data warehouse allows an e-business to consolidate its Web site data with data collected from other stores, including OLTP databases and external data sources. This combined data can undergo transformation, cleansing, and enhancement before it is stored in the data warehouse. The data warehouse then becomes the primary source for analysis of all company data.
The consolidation offered by the data warehouse facilitates the analysis and leveraging of the data from a common pool. Any of these techniques can be used to conduct analysis in the data warehouse.
Data warehouses can simplify the process of consolidating, analyzing, and leveraging a growing pool of company information, and good data warehousing systems are a strategic asset for successful e-businesses.