Data Mining – Understanding Your Data Leads To Profits
March 5th, 2010Information captured by your online business plays two vital roles. First, it allows you to directly service the needs of your customers. Second, it empowers you to make decisions that will allow your business to grow. Of course, some may argue that a growing, thriving business benefits your customers and a well serviced customer base allows your business to grow. All true. But for the sake of this consideration, we will be segregating your data into these two classes: transaction processing and decision support. Let’s explore them individually and see why a clear understanding of each will help you provide the best experience possible for you visitors and afford you the finest insights into your business’s future.
Transaction Processing
When a customer does something on your website worthy of recording or capturing, that is typically considered a transaction.
· The customer signed up
· The customer wants a news letter
· The customer purchased something
· The customer adjusted his order
· An item was shipped
In all of these cases, an event took place that has related information worthy of your attention and important enough to be captured. It may prove beneficial to you as a business owner who will make a decision, or it may prove helpful in servicing a customer; perhaps someone wants to know when the widget was shipped, for example. These are transactions, and depending on the amount of business you do, there could be a large number of them each day, perhaps many a minute. Small, discrete pieces of data inserted into the database. This activity represents where your business is at this very moment; the invoices that need to be cut, the product that needs to be shipped and the customers that deserve an email thanking them for their business. This is live, “hot” data and it is likely that it will be referenced in the near future. This is compared to “cold” data or decision support data.
Decision Support:
There is a difference between reordering a case of widgets because you are about to run out and determining whether you will be carrying widgets next year.
· How well did this item sell compared to other items in the same category?
· How is my business trending?
· What are total sales to customers based on region?
· How do those sales compare to last year?
A review of these questions reveals that the data is going to be used to create a different view of the business; one that is a bit less specific in nature. Rather than concerning ourselves with an individual sale, a transaction, we are concerned with a number of them. Data is group together so that we can step back and see the business from a different perspective. It’s analogous to looking at a city as an urban developer as opposed to a shop owner. Both have their place, but one is clearly more interested in the big picture and the decisions that affect the long term future of the community.
Any business that is to succeed needs to balance both needs. Customers need to be serviced and their immediate questions and concerns need to be met. Invoices need to be cut and product needs to be reordered. In addition, the overall course of the business needs to be addressed. Are you on track to meet your goals? What are your developing markets and which ones are no longer of interest to you?
When you logically separate your data and see the roles that it plays, then it may be time to consider physically separating it into separate entities dedicated to each task.
I will address the potential benefits of aggregating your data and customizing your storage to meet your specific business needs.
