Victoria's Secret Direct racked up almost $1 billion in net sales last year, with the Internet channel contributing a considerable portion (including double-digit growth).

Director New Media Marketing Steve Kahn says these online sales are "...really about customers choosing the channel that's most appropriate at the time and place they're interacting."

A successful multichannel business is one that can be there with the right channel at the right time to match customer behavior.

But this requires an evolution beyond what Kahn calls the first wave of Web metrics, such as "...how many people came to my site? How many unique visitors did I have? What were the basic paths they were using?"

He explains, "The next wave will be a more sophisticated approach that will connect the information about the customer's experience in the brand, and that includes contacts in stores, contacts from catalog and interactions online. Businesses will need to be smart enough to build the right rules and tests to make sure we're providing the right experience for the client and the business."

How? Kahn suggests three steps are needed to take your multichannel business to the next level of measurement:

Step #1. Understand the levels of behavioral targeting

The current cutting edge is personalizing Web site presentation based on the past behavior of the visitor in question.

At the most basic level, "there are different data sets that you can work from; there's the data from your past history, what you purchased, what you looked at, etc."

The next measurement level is an experience that includes customization based on what a customer's actually doing in a current session, "...because sometimes that will trump what they've done before."

The third, most advanced level is then, "...to combine both the logic of what they've bought in the past and what they're doing now, and actually build some rules around that in order to best understand what that person might be looking for, and to better estimate what other products or services they might have an affinity for."

"That's the utopia."

Step #2. Measure (and understand) quantitative versus qualitative

Kahn notes that another issue retailers need to recognize more is that Web metrics tell you what people did, but not why.

"You have Web metrics that can tell you the path that the people who land on a certain page will go to. That's good, quantifiable information. Why they did it, you might have no idea - you can only make assumptions based on the knowledge of your site, brand, and customers."

The challenge, then, is to use the quantitative information to understand the qualitative reasoning behind visitor choices, thus highlighting the value of analysis and testing.

"Having the metrics doesn't give you the answer, it permits you to ask more intelligent questions... if people are going from one place to another and you never intended that to happen, then you need to sit back and say, well, why would they do that? For key business functions, this is a great opportunity to utilize user testing and panels to connect the dots."

Kahn describes this as the finer art of Internet analysis, "...combining the quantitative with the qualitative to come up with a better experience."

Step #3. Define your measurement goals

Kahn also suggests that retailers will need to better define and standardize (for themselves) those metrics that measure success. "You have to be very specific as to what you're looking for... every business has to figure out what is the key metric that they're going to judge their success by."

One key metric Kahn's team uses is account profitability by traffic source. There have been some surprises, such as the typo factor. Kahn says in total more people spell the Victoria's Secret brand name wrong in the URL than spell it right... "though the people who spell it right are better customers!"

What's the biggest constraint to metrics evolution? Kahn says staffing is going to continue to be a huge challenge.

"As more people want Web sites that function well and are effective for their brands, there's a smaller group of people who actually have experience marketing and running things in the Web medium."

"Across the industry a lot of people are finding that same kind of challenge in terms of getting people with a deeper skill set who are more online based. There's not a lot of people with experience out there, because it hasn't been around that long."

So for the time being, a staffing crunch may be the most significant factor holding back great Web analytics for many companies.

Note: Victoria's Secret is a member of Shop.org, a forum for retailing online executives to share information, lessons-learned, new perspectives, insights and intelligence. More info at http://www.shop.org