Today marks the debut of a new featured columnist for eCRMGuide. Arthur O'Connor is a senior manager in the financial services practice of KPMG Consulting specializing in customer-facing strategy as well as related architectural and organizational issues. An accomplished author, speaker and consultant, Arthur is one the country's leading experts in customer relationship management (CRM) and eCRM. In his first column, he'll discuss what he considers some of the key trends in the eCRM field.

For my first two columns, I thought I'd hit some of the key trends in the eCRM field -- trends that may well figure prominently in our industry, and our lives -- as the year unfolds.

With so many new developments in eCRM, it's sometimes tough to tell the difference between false starts and lost causes. By my reckoning, I see about five major trends that directly influence how eCRM is being designed and deployed in corporate America. I'll cover the first two here, and two weeks from now, I'll follow up with the remaining three.

Time will tell how accurate these predictions turn out to be (I expect e-mails proving me wrong will commence by summer). But here goes:

1. We'll learn to better use the Internet as a CRM tool

This is a pretty safe bet, because we still have a lot to learn. Among the great lessons learned from the dot com recession are that, unless you're selling porn, the Internet is not a great sales channel, not a great advertising medium, nor even great a direct marketing tool. It is however, a potentially powerful and cost-effective customer intelligence-gathering and relationship-building channel, which few companies have fully exploited.

With a better understanding of the new medium's strengths and weaknesses (e.g., the web is better at selling known products like books, than selling products that require trial or personal examination, like blue jeans), businesses will design on-line customer-facing strategies that are more consistent with on-line human behavior.

The key benefit of the web as a commercial medium is that people are willing to do some things over the Internet, buy books and CD's; enter their own orders or access information, that they were less willing to do with other technologies (writing a letter, using the phone or a fax machine). It is this willingness of consumers to cost-effectively "self-serve" that enables on-line companies to achieve economic benefits (akin to what ATM's achieved for consumer banks over the past two decades).

The on-line experience as it relates to consumer behavior has limitations. Most consumers still prefer to take out a mortgage at a local branch. While some banking sites are great at attracting on-line visitors, they are notoriously bad at facilitating transactions. As we learned last year, most businesses don't want to risk procuring goods from anonymous suppliers on B2B exchanges. They prefer to do business with people they know and trust.

Drawing upon these valuable lessons, businesses will begin to use the Internet channel more effectively, as part of their total customer-facing strategy. They will also learn to use it more discretely.

The majority of commercial sites today talk to and sell to customers, instead of listening to and serving them. Instead of focusing on click-through rates and transactions, I believe (and hope) businesses will begin to appreciate on-line customer interactions as opportunities to extend a relationship (not necessarily close a sale).

2. We'll get smarter about how to use client data

In the 1990's, the push was to get in front of the customer and "own the relationship." Most companies realized, as witnessed by the huge investment in sales force automation and customer relationship management suites over the past decade, he who owns the relationship owns just about everything. We're not just talking about the lion's share profits along the supplier/producer/wholesaler/retailer "value chain. "More importantly, we're talking about the value of knowledge about customer needs and buying habits.

Why has customer data become so important? Because in the new global economy of white-hot competition and frenetic change the only real sustainable competitive advan

tage in attracting and retaining customers will be to know them individually and better than anyone else does, and to be able to offer them what they want, when they want it. This means providing them financial, emotional and other behavioral incentives to stay, and offering personalized, valued services and add-ons.

While most organizations have the ability to collect information, they don't have a coherent and effective strategy to analyze and process this tremendous amount of data into actionable information.

The current wisdom in both the on-line and off-line worlds is that, with storage memory costs continually falling, if you don't know what information to keep, keep it all. Such economics have forced businesses into a data-rich, information-poor scenario. Corporations that have undertaken data warehousing projects have learned the hard way just how complicated and difficult it can be to decide what information to look at, how to analyze it, and what to do with the results.

Despite accumulating huge amounts of data, most businesses have quickly come to realize (or at least strongly suspect) that all customer data aren't created equal. While the science of data analysis is not new, the art of discovering what customer information has predictive value is. Most techniques employed today are still based on traditional, sales-transaction-oriented metrics, such as RFM analysis (recency, frequency and money) analysis, which tracks customers by how much, recently and frequently they purchased your products, techniques which have been around for decades.

For better and for worse, the Internet has its own set of metrics. The good news is that every move a site visitor or on-line customer makes can be automatically captured in a web server log file. Because this information is already electronic, it is relatively cheap and easy to analyze and manipulate.

The bad news is that many companies get carried away with the beauty of the reporting. Some companies spend tens of thousands of dollars on detailed reports of on-line visitor behavior with absolutely no clue how to recover this cost in the form of smarter marketing, more targeted sales, or more informed product development. Many businesses are waking up to the fact that just because it is cheaper to track every move a visitor makes on your site doesn't necessarily mean it's worth doing. While some click stream trends are predictive (for example, increased frequency of a unique visitor to a site indicates likelihood to make an on-line purchase, and collaborative filtering can be an effective sales suggestion engine), on-line behavior modeling is far from mature.

Is there a common theme in these eCRM industry trends and predictions? Well, it could go something like this: despite its youth, eCRM has done a lot of growing up lately. Despite snazzy lingo, fiendishly complex technology, and the continued novelty of the Internet, in many ways, it's back to the future: managing by the numbers.

Compared to the gut-wrenching hoopla investors, technologists, sales people and consultants have suffered through, maybe that's not such a bad idea.

Tune in two weeks from now for the next three trends in eCRM: multi-channel integration, impact on the supply chain, and implementation.

Arthur O'Connor is a senior manager in the financial services practice of KPMG Consulting specializing in customer-facing strategy as well as related architectural and organizational issues. An accomplished author, speaker and consultant, Arthur is one the country's leading experts in customer relationship management (CRM) and eCRM.