eDave: One of the truths that the dot com phenomenon has finally reaffirmed is that people choose to interact with companies for one of three reasons: to research a product or service, to purchase a product or service, or to get customer support for that product or service. Those have always been the three reasons people interact with companies-all the way back to the earliest stores. In the early days of the dotcom craze, because the Web was a completely new channel for both communication and conducting transactions, some people over-focused on e as a selling channel.

Any channel can serve to support those three types of interactions. Most people think of the e-channel as a Web-connected PC. But looking into the future, I think "e-channel" will also mean e-enabled ATMs, kiosks, or other wireless devices that support these 3 interactions between the customer and the company through Web-like interfaces, even if that mechanism doesn't look like today's browser-based PC. What's most important for companies, though, is to remember that as these e-channels proliferate and evolve, the reasons customers choose to interact via these channels-the basic motivations-will remain the same three that existed hundreds of years ago at the simplest stores. Research, purchase, service.

eDave: I don't think so. Customers have always been motivated by convenience, and the e-channels are very convenient. Price can be a factor, but often convenience will trump price. Customers are becoming more and more accustomed to relying on the Web and e-mail in their daily and business lives. They will increasingly expect the convenience of communicating and interacting with companies via e-channels. Customer-focused companies will understand that today's primary business driver is convenience.... which means, in a nutshell, ignore e-channel marketing at your own peril!

The facts back this up. One study by Jupiter Research recently showed that most consumer behavior is cross-channel. For example 68% of people have researched products online, then purchased in the physical store. Another 54% researched products in stores, and then bought online. 47% have researched using catalogues, then bought online. So the importance of the e-channels for research and service aspects can be as important to a company as the facilitation of sales. Another "e factoid" reported by Peppers and Rogers recently is that 10% of all major appliances purchased in the stores at Sears occurred after people researched appliances first on the Web, and that those sales totaled $500 million.

eDave: An ideal implementation would focus on the interactive nature of the e-channels, so that your marketing team can gather more clues and try more things out to understand and learn about customers in a shorter amount of time. Marketing is the design and execution of campaigns to influence people's behavior, broken down into these steps: segmenting customers, developing appropriate messaging, sending messages through the appropriate channels, and measuring the effectiveness of the campaigns based on customer feedback or sales metrics. Ideal e-channel marketing is not separated from marketing at all, but rather is integrated into the overall marketing strategy.

However, what's new-and exciting-about e-channel marketing is that it enables marketers to segment based on actual behavior, not just demographics or psychographics. You can't track actual behavior in a brick-and-mortar store-at least, not without stalking customers. But through e-channels, tracking and analyzing Web clicks allows companies to do a better job on properly segmenting customers based on clues about what they really are interested in. You don't have to guess that someone might be interested in a new outdoor barbecue grill if you know that they visited your Web site twice in the last week and spent 22.5 minutes looking at grills.

e-channel marketing also allows businesses to improve on another fundamental of marketing-messaging-by very quickly testing marketing messages in a way that simply wasn't possible before the advent of e-channels. For example, a company can test 10 variations of its buy-a-grill messages with subsets of various customer segments to analyze the variations' effectiveness for various segments, then fine-tune based on that feedback, before launching messages to broader segments or across other channels. This fine-tuning and timing of messages just isn't possible through other channels with mass advertising. Putting the grill ad on the home page the next time that visitor comes back is an obvious way to make your Web site more relevant to the customer. Purchase rates for well-tested campaigns can be as high as 25%. Bad campaigns get 0.7% conversion rates.

Channel mix is the next area where e can make a difference. Not only will more channels like bank ATMs be Web-enabled, but even TV will become interactive. Work is underway to explore how cable advertising can be fine-tuned to fit the interests and needs of the person watching the cable show, so that if you're over fifty, perhaps the travel advertisements you will see are for relaxing, spa-oriented destinations, but if you're college-age, travel advertisements will focus on the popular spring-break destinations. e-channels can also reduce costs. For example, in many studies, there is 5-fold cost reduction by moving service interactions from call centers to the Web site. Surprisingly, it turns out that when people interact with your Web site instead of a human, they often rate your company higher in customer service scores.

Finally, measuring and analyzing campaign data, including cross-channel behaviors, can help the ideal marketing group glean more insights per unit time than competitors. For example, on a typical retail Web site, analysis of Web browsing patterns may show that customers research and purchase in five ways on your Web: via category, brand, trends or fashions, things-on-sale, and size. Price sensitivity may not be the same across these 5 browsing patterns, so a smart marketing group could take advantage of that insight.

In summary, a good e-channel marketing effort takes advantage of these 4 aspects of the Web. Behavioral insights result in better segments. Testing of messaging increases its effectiveness with those segments. The Web can be a cost-effective part of multi-channel campaigns. Metrics help marketing people work smarter.

eDave: My two pet peeves are Web sites with annoying layouts and those with irrelevant banner advertising.

Some site layouts are hard to use and could benefit from some rigorous customer use analysis - what do people do frequently on your Web site, and how can you help people do what they want to do easily? Remember that most people use this channel because of its convenience.

The other problem is sites with too much advertising, especially if it is not well-targeted. Instead of concentrating on more relevant advertising, some people create even more visually distracting ads - things that blink or scroll, or won't go away easily. I've been spending a lot of time lately thinking about how companies could avoid what I call "clutter advertising." If you collect Web clues, your marketing campaigns (and for that matter, your physical mail campaigns) can be much more focused and relevant. If you do traditional "carpet bomb" advertising, you not only waste customers' time while annoying them but you use up financial and other resources that your marketing departments could be using on more effective communications.

eDave: We're seeing four stages. Most companies are only in the first or second stage. The first is a one-size-fits-all Web site. The company creates its presence on the Web and provides product or service catalogues with the ability to order. The second is allowing people to customize the Web site layout to make it more convenient to them --- the "My Web Site" angle. The third involves building a dynamic Web site in which the company personalizes information for people based on their previous purchases with the company. However, this is still "push"-based marketing. The final stage is a truly smart Web site that incorporates all interactions with the company to hold smart dialogues, across all 3 customer behaviors - researching, purchasing, and service. This last stage requires the integration of information across all the customer touchpoints so that a dialogue can pick up on the Web where it left off with the call center agent or in the store. Sometimes customers initiate the dialogue; other times the company begins the conversation.

eDave: Companies must think holistically-that is, they must not separate e-channel marketing from other marketing efforts. Leaders must recognize the treasure trove of customer insights that the Web can generate, and use the information in an enterprise data warehouse to drive smart dialogues on the Web channel. In this sense, Teradata users have a leg up on their competition because they already have developed some discipline about gathering corporate information in one place. Adding the Web data to that, and then deriving complete customer insights across all channels, stirring in the opportunities to do experiments on the Web - that's where I think the action will be. We've started documenting these Best Practices so current Teradata users can learn what others are doing.