Heather Kaminetsky, director of Internet marketing for Barneys New York, has been asked more than once how she knew a particular person would like to receive more information about a particular product.

It’s not that she’s secretly been spying on the buying habits of others – or that she has some innate ability to perceive likes and dislikes. It’s more that she has a cutting-edge tool at her disposal that can, in essence, predict what customers want, when they want it – and at what price point.

“It sounds like the Holy Grail, doesn’t it?” asks Sheldon Gilbert, founder and CEO of Proclivity Systems, which landed Barneys as the first client for its Proclivity Mail technology. “A lot of people have wanted to get there for a long time. But it’s hard, particularly online. The biggest problem we’re seeing is people trying to use old-world, offline paradigms to do intelligent selling online.”

In the past, store sales clerks would observe customer behavior first-hand. “But as we’ve grown larger and developed the online shopping mall, the cash register as indicator of interest has become far too myopic,” Gilbert says. “You interact with many, many other items, but only a fraction of them do you actually buy.”

Proclivity Mail provides a way for vendors to not only track what shoppers are interacting with online, but also to translate that information into targeted e-mails with higher open, click-through and purchase rates.

Once the data is gathered, probabilities are generated for the number of customers apt to bite through “advanced econometrics modeling,” and marketing campaigns can be more strategically focused to garner success. (Barneys will not divulge its own statistics, but Gilbert says some Proclivity clients have experienced 13-fold jumps in order conversion, 15 to 26 percent increases in revenue and 600 percent increases in campaign productivity.)

“E-mail companies all have segmenting, but nobody has it like this,” Kaminetsky says. “I’ve had [colleagues at Barneys receive] an e-mail and say, ‘How did you know I’d be interested in this beauty product?’ And then they forward it to a friend as say, ‘Did you get this e-mail?’ and the answer will be ‘No.’”

Lowering the “nagging index”
Though Proclivity has information on various customers, it doesn’t keep e-mail addresses and, therefore, doesn’t “know” who those customers are: Batch lists are generated according to session IDs, Kaminetsky says.

“I don’t think people would be shocked to know that their behavior is tracked in that way, and I don’t think they’d see it as bad,” she says. “We just want to know what customers are looking for, so we don’t waste their time. We realize that time is valuable.”

Sending targeted e-mails rather than blanket blasts “gives retailers the ability to lower the nagging index,” Gilbert says. It also allows them to save some money, since they’re charged on a per-thousand-e-mails-sent basis.

“This can give you the ability to engender some loyalty,” he says. “People who receive e-mails at a lower frequency with a higher relevancy are much more likely to become buyers. This is a way of tuning into their interests.”

Gilbert says he approached Kaminetsky – an associate from the days when he ran the business intelligence practice at Bluefly.com – because he knew that the company was in the process of growing its online presence. (Though Barneys has gained great success as a fashion-forward luxury goods retailer, its e-commerce website is only two years old.)

“One of the first things I wanted to help them do was to market to their existing set of customers as well as to their new customers,” Gilbert says. “And with a new system, it’s easy to make sure the right framework is in place from the start.”

Proclivity performed a pilot run last summer, and Kaminetsky was impressed. “There’s definitely a large opportunity out there,” she says.

Online advantage
A woman might go into a store to buy a handbag, she says, and may happen to browse through men’s shirts for a boyfriend or husband while she’s there. “In the store, no one may notice that the woman has done that,” she says. “But online, we know.” And that behavior can be followed up with a targeted e-mail when there’s an item that may be of interest not only in the handbag department, but also in men’s shirts.

Kaminetsky has asked Proclivity to run periodic tests to monitor the system’s effectiveness, and the results have been consistent.

“We’ll take two creatives, something like handbags and men’s shoes,” she says. “And we’ll say, on this day, give me everyone who’s interested in women’s handbags, and that’s list one, and then men’s shoes, that’s list two.

“The handbag e-mail will go to those interested in handbags, but for a small group, we’ll actually send the one about men’s shoes. We want to prove that this really works, and the only way to do that is with negative control groups. And we find that when we send the men’s shoes e-mail to people in the women’s handbag group, the open rates, the click-throughs and the sales go down.”

Barneys is sending out between four and six e-mails weekly, and its customers are “responding well,” Kaminetsky says. Future plans include working with Proclivity on cross-merchandising. “We do plan to use it more and more,” says Kaminetsky, who believes it to be the wave of the future in terms of e-mail marketing.

“The beautiful thing is that all of this data already exists,” Gilbert says. “All of these systems are collecting vast terabytes of data all the time. The problem is that it’s not being used for a number of reasons. It may be that the volume is too overwhelming, or that it’s too complex. So we developed this engine that will take all that in and recalculate the relationship with every customer and the products they come across, and predict whether they’re going to buy them.”

With traditional engines, there’s a lot of data for the customer, Kaminetsky says, “but not so much for the subscriber. Tons of data goes into databases every day, and there’s tons of information about potential customers, but you don’t know about it unless you really dig into the database.”

Proclivity “is not trying to sell you information about your customers that you already know,” she says, “but information about the subscribers and what their interests are.”