Data Mining and CRM
We came across this article at it.toolbox.com/blogs/insidecrm/data-mining-and-crm-76516
Data mining is the process of using statistical algorithms to collect customer information and which can be used to execute a more precise marketing strategy.
As such, data mining can be an important adjunct to your CRM efforts. In fact some CRM programs now include simple data mining tools as part of the package.
The first thing you need to do to make effective use of data mining is to decide on a data mining strategy. What are you trying to find out? Like many things in CRM, data mining can accomplish many different goals, so it's important to decide what kinds of insights you're after and what you intend to do with the information.
Choosing the right tools is important for a successful data mining effort. Available tools range from inexpensive to pricey and from simple, easy-to-use software to things designed for use by data scientists. It's important that you match your company's IT expertise with the tool or tools you use.
The next step is to slice and dice your customer base into sub-lists. Each of the members of these lists should have more in common with the other members than they do with the list as a whole. You can do this by amount of sales, demographics like age, zip code or anything else you can think of. Make lots of lists because they will help you extract usable information from your database.
Using the data mining tools, examine the lists for commonalities. Try to find what your customers and the sub-segments have in common. What behaviors do they share?
You can also use data mining to improve relationships with vendors and other businesses. Often predicting vendor behavior is at least as important to business success as predicting customer behavior.
This is the information you can use to predict customer or prospect behavior. That information can be used to formulate strategies to increase your sales.
One of the most important results from successful data mining is to predict who is likely to respond positively to various appeals. This can increase the number of hits you get from your campaigns and reduce the number of duds that aren't likely to purchase as a result of this campaign.
Properly done, these predictions are surprisingly accurate. However if you're building a major campaign around them it is best to test market the campaign first. The data may not lie, but it can be misinterpreted and that can lead to major mistakes.
Feedback is also an important element of data mining. Check your results against each other and common sense. Also monitor your strategies and their results as you run campaigns. Use the feedback to check that you are getting the results you expect and if not, figure out why not.
Finally, data mining is a supplement to customer contact or other forms of data analysis. Properly applied it is a very important tool. But it's important to recognize the limits and to supplement data mining with other kinds of information.
About the Author
Rick Cook has been involved with computers since the days of punched cards and magnetic drum memories. He has written hundreds of articles on computers and related technology as well as a series of fantasy novels full of bad computer jokes.