Customer data mining and personalization

In this tip, find out how to practice customer data mining to personalize customer marketing interactions.

The single thing most people lack today is time. If an e-business can reduce the time required to search through

a catalog or directory by anticipating what product or service visitors are likely to want, those visitors are not only going to be very grateful for that level of service but much more likely to buy something as well. To deliver that level of eCRM personalization, an e-business must provide instant and relevant product offers. The best way to do that is to employ the power of data mining. Outsourcing Personalization However, many companies see this type of data mining-driven eCRM personalization as quite expensive and quite daunting. But the web has changed all that. One of the e-business "killer apps" of the Web is the ability to outsource this type of data-driven personalization to experts at modeling, customer profiling and predicting consumer behavior and product preferences. A data mining driven application service provider (ASP) can provide personalization to an e-business as a service over the Web for a fraction of the cost of an eCRM. The Web really does alter many things. Porsche, for instance, does not build the Boxster. They design it and arrange for its construction by outsiders with final assembly of the sports car in Finland from where it is shipped directly to dealers. This would not be possible without the Web. Today's successful e-businesses recognize the value of outsourcing, too. Such tasks as hosting, e-mail management and credit card processing are commonly outsourced. But, until recently, few e-businesses were ready to contract out one of the most difficult task of e-commerce: personalization. Today, personalization can be outsourced to a data mining ASP without the need to buy any software, hardware or hire additional IT staff. Providing personalization involves aggregating clickstream data, enhancing it with demographics and mining it all in order to extract predictive models for anticipating visitor behavior and consumer preferences and then making real time offers. This work is not for the inexperienced. To do it in-house is very expensive, time consuming and difficult to execute, if at all. However, the right data mining ASP can provide the service using experienced professionals with guaranteed ROI. Privacy and Segmentation It is very important to note here that care however must be taken in the use of certain Internet mechanisms like cookies and Web bugs (invisible graphics) to increase personalization and ease of use. These hidden techniques track and collect individual browsing actions without visitors' consent or knowledge. Personalization is provided but at a cost of violated customer privacy. There are ways of providing personalization, while protecting visitors' privacy and obtaining profits for an e-business. One way is by using offline demographics. As visitors interact with an e-business, the site might prompt them for their ZIP code, in order to deliver local news, weather, and coupons or to locate a nearby store. That simple ZIP code can be used to segment customers by data mining techniques in order to discover product affinities on the basis of neighborhood demographics. These are traditional techniques that have been perfected by database marketers for years. It involves a data mining process known as micro-segmentation. The difference here of course is that the offers are made in real-time, rather than through traditional direct mail. However, more importantly because a ZIP code is used, the privacy and anonymity of the visitor is guaranteed, while personalization is still provided by that e-business. ROI and Costs The majority of companies that invest in eCRM have little or no idea on how to estimate their return on investment (ROI). They have no measurements by which to compute costs against profits. Related costs should include not only the software, its installation and deployment of the eCRM system but also the impact of delays in existing business processes. In addition, the cost of external and internal labor and the training of personnel, which at times may cost more than the software and hardware, must be covered. A Forrester report "CRM: At What Cost?" recently observed, "Many companies have made no formal evaluation of their eCRM program, and as much as 20 percent of a typical eCRM implementation is wasted due to overlapping software." This identifies an often overlooked cost, that of redundant IT systems. Once all of these costs, direct and indirect, are considered, acceptable ROI on a CRM installation becomes almost impossible. In this light, outsourcing personalization from an ASP begins to look highly appealing, smart and cost effective. Service Over Software The end of software as we know it is upon us. There is an undeniable revolution away from software to services involving a new delivery and management model. The new way of buying business solutions is incrementally - via a subscription model over the Web. Why buy when you can rent an application and avoid all the headaches of operation and maintenance? Personalization is a very complex and expensive application involving several processes, such as data preparation, demographic enhancement, as well as the capability of creating predictive models and leveraging them in real time, in order to anticipate and influence customer behavior. If there is one application, which is ideal for outsourcing, it is data mining, personalization and eCRM. Any e-business can be like Porsche, savvy enough to know what it wants and smart enough to know when to have experts build it for you. Jesus Mena is the CEO of WebMiner, Inc. and the author of Data Mining Your Website and WebMining for Profit.

This was first published in October 2001

Dig deeper on Data quality management in CRM

Pro+

Features

Enjoy the benefits of Pro+ membership, learn more and join.

0 comments

Oldest 

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to:

SearchBusinessAnalytics

SearchDataManagement

SearchSAP

SearchOracle

SearchAWS

SearchContentManagement

Close