Individually explosive... The Web has changed where and how companies and individuals get information, make transactions, communicate and conduct business:

  • Sales via the Web have grown rapidly. Sales of books, computers and travel, in particular, have experienced triple-digit percentage growth for some firms' Web channels. Services such as electronic banking are also growing and changing the marketplace.
  • Companies are using the Web to communicate with both customers and suppliers, gaining substantial reductions in lead times and significant improvements in inventory management.
  • Communications are faster and more convenient on the Web, with uses ranging from providing shipper information on package delivery to providing a company's managers across the globe with the latest internal financial results.

Data mining also provides companies with competitive advantages in optimizing their use of information. Data mining is used to:

  • Manage customer relationships by analyzing customer segments, target marketing and promotion effectiveness, customer profitability, customer lifetime value, customer loyalty and retention, and customer acquisition effectiveness.
  • Position products by product affinity analysis that shows opportunities for cross-selling, upselling and strategic product bundling.
  • Manage human relations by gaining an analytical understanding of the patterns in employee retention.
  • Enable financial management through analytical fraud detection, claims reduction, detection of high cost to serve orders or customers, risk scoring, credit scoring, audit targeting and enforcement targeting.

Combined, they have powerful synergies... The Web is a powerful information disseminator, allowing for individualized messages to be delivered cost-effectively. Data mining and related analyses are powerful information developers, allowing for strategic actions based on knowledge. The combination is a natural.

In particular, there are three areas where data mining and the Web combine to make each other more powerful: customer service, target marketing and internal business communications.

Customer Service Customer segmentation can help analyze how different types of customers use a particular service over the Web. Companies can then tailor services to better meet the demands of those various groups of customers. Even earlier in the process, data mining models can be used to predict which customers could be switched over to the Web, often a more cost-effective channel for multi-channel companies. When a new customer is gained, data mining of early transaction patterns of existing customers can predict how valuable this new customer is likely to become. Interactions with the new customer can then be tailored to create differentiated service based on future value.

Data mining can also improve the process of ensuring that customer orders can be met on time. A manufacturer can use statistical modeling with its customers and suppliers over the Web to ensure it has accurate forecasts of customer demand and supplier inventory.

Target Marketing Once a relationship is established with a customer who does business electronically, there is an opportunity to keep that customer well informed of products or services. But what information should be provided in hopes of capturing a greater share of that customer's wallet?

Describing all products or services to every Web customer means not only missing the power of tailoring the message to the customer, but also demanding that the customer spend time looking at volumes of material. Time is valuable to Web customers and anything that wastes it will be, at best, ignored.

A succinct message about a single or a few products/services is more effective. Tailoring the choice of which product/service is communicated to which customer is vital. The more meaningful the information, the more likely they are to respond to it, and the more likely they are to regard the service as positive. How does one choose which product/ service is contained in the message given to a particular customer?

One might think that business rules would be an effective approach to targeting a product message to a customer. Examples are:

  • Pitch a new history book to a customer if the last book purchased was in the same genre.
  • Send a buy-one, get-one-free promotion to the top 100 customers.
  • Inform customers in wealthy ZIP codes about a luxury product.

However, such rule-based actions can frequently be ineffective. What if the customer bought the history book to give as a Father's Day present and, personally, has no interest in history? What if the top 100 customers already feel they are purchasing enough of the product and don't change their buying levels due to the promotion? What if the wealthy residents of target ZIP codes formed their preferences when they were still on their way up, and promotional information on a luxury product comes too late to influence them?
Rule-based actions can, in some cases, even backfire. They are often simple and obvious, especially to the savvy customer who uses the Web. As such, they appear artificial--an attempt to show your affinity to the customer when, in reality, you haven't done your homework. In the end, they may actually annoy the customer.

Here is where data mining can help. Based on analyzing data to identify which types of products appeal to certain specific customers across your customer base, you can identify the product message that has the greatest likelihood of being effective--not only in terms of increased sales, but also in terms of strengthening the relationship you have with the customer by truly showing you can tailor services and product offerings to their needs.

Internal BusinessData mining is frequently done centrally or regionally. However, those on the front lines need to have the knowledge gained through data mining. They will sell to and service customers, manage inventory, supervise employees and work to correct and prevent loss. The Web can be an effective means of allowing operational personnel and suppliers around the world to quickly gain the information needed to effectively do their jobs. Information derived from data mining can be communicated to operational employees in several forms:

  • An algorithm for scoring. For example, provide employees with an algorithm to take the characteristics of a transaction and use them to calculate an estimate of the likelihood that it might be a fraudulent transaction.
  • A score for a particular customer, employee or transaction. For example, transactions can be scored overnight and those that are potentially fraudulent flagged for employees by morning.
  • A recommended action associated with a particular customer, employee or transaction. For example, transactions can be scored overnight, and an action item such as "obtain additional validation" communicated to employees by morning.

Four Tips DELIVER THE FUNDAMENTALS: Don't let the incorporation of knowledge gained from data mining get in the way of quick, hassle-free communications--the reason you went to the Web in the first place. Also, keep in mind that Web customers are different from non-Web customers; therefore, any data mining results derived from analyzing an entire customer base may not be applicable to a Web-customer base.

RESPECT YOUR CUSTOMERS: If using data mining results to enhance interaction with customers over the Web, remember that the end goal is to maintain and strengthen the relationship with the customer. Therefore, avoid creating time-wasters, hassles and perceived threats to their privacy. For example, if the customer has signed up for the Direct Marketing Association's mailing or telemarketing preference list, he or she has indicated a strong desire for privacy, a strong desire to cut down on clutter, or both. Use this information, along with any complaint information, to flag customers who prefer to tell what they want rather than to receive information from you.

DON'T SKIMP ON DESIGN: Those delivering written or televised messages to customers spend a great deal of time ensuring they have a design that promotes their effectiveness. The same standards apply for messages delivered over the Web. Involve an expert in Web design to incorporate your tailored messages. These experts can help address areas such as clutter, color and placement to ensure your message is noticed and that the look of the message enhances its effectiveness. Also, make sure that your Web image and product offering are in sync with your other media images and customer channels. A direct mail company that can provide in-stock inventory information to customers over the phone should be able to provide the same service to its Web customers. A company with a sophisticated print campaign should be equally sophisticated on its Web site.

MAKE EFFECTIVE USE OF TECHNOLOGY: Time is of the essence when communicating via the Web. Therefore, the shorter your cycle time in identifying and acting on a data mining result, the better. You must have a sound data repository with high-quality data, accessible for analysis purposes, and with the range of data required for effective data mining. A data warehouse fills this need. Data mining can then be executed against samples from the warehouse, and the resulting algorithms used to score all members of the warehouse, saving scores in the warehouse for ease of use. The scores can then be pushed out to where they are needed: on the Web for employees to use, into a campaign management tool to design and execute a Web-based campaign or translated into actions carried out by a Web application. There are various products on the market that enable the personalization of communications via the Web. Some help with general target marketing, others facilitate personalized service and still others allow for complete, one-to-one communications, transactions, services and support to match the needs of employees or customers. All are made more powerful by the use of data mining to determine which message is most appropriate for which Web user.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access