Wilsons Leather: Turning web data into revenue
“We can track customer behavior on our web sites on a 24/7 basis, but why do we feel so overloaded by the data we can now collect?” asked Rose Hamilton in her presentation last week at eTail 2002. “It’s like a lead weight, a bowling ball around our necks.”
As e-commerce marketing manager at Wilsons The Leather Experts Inc., Hamilton spearheads an aggressive program for the leather apparel and accessories retailer that she says is designed to cut through the mountains of web-generated shopping data to “identify the nuggets of information that drive online revenue.” Most web retailers, she argued, are so overwhelmed with data that they don’t bother to wade through it looking for those nuggets, preferring instead to stick with simplistic measurements such as overall site traffic.
By comparison, the web merchants at Wilsons Leather treat the data overload as a marketing asset and have developed comprehensive and detailed methods for analyzing the data with an eye on improving web sales and conversion rates. The analysis covers all aspects of web activity, from the effectiveness of outside as well as internal search engines, to the click behavior of shoppers on the site, to the effectiveness of its e-mail and other direct marketing programs. And as soon as each analysis of the metrics is completed, the web site is changed. “We push the limits on how quickly we want to make changes to the set whenever we look at a new set of numbers,” Hamilton said.
For example, the company’s detailed analysis of the performance of outside search engines is routinely summarized in a “search optimization report,” which shows which keywords, search categories and outside search engines are producing the highest order volume and highest ratio of orders to searches. By applying that information to keyword acquisitions and other aspects of its use of outside search engines, Wilson Leather in the last year increased its sales from search third-party engines by 24%.
Similar improvements have been achieved through detailed analysis of the company’s e-mail marketing program, the metrics for which are regularly summarized in an “e-mail campaign report.” That report analyzes how recipients of e-mails promoting the site respond to different marketing messages and different product promotions; how they browse the site and when and where they abandon it, how many times they return to the site before they make their first purchase and how often they return for repeat purchases. “It has totally changed the way we measure our e-mail programs,” Hamilton said. “What we measure now is the total impact of e-mail over the long-term, not just whether e-mail recipients came to the site to buy or not.”
Wilsons Leather takes the same detailed approach to analyzing each web visitor’s behavior at the company’s web site. It determined, for example, that an annual motorcycle giveaway designed to enlist registrants for the web site produced lots of registrants and e-mail addresses but very little sales. “People came to the site to register for the bike drawing and then they left,” said Hamilton. The company attacked the problem simply by positioning merchandising promotions on the home page right next to the registration for the motorcycle drawing. The result: this year’s motorcycle contest resulted in a 27% increase in sales attributable to it.
And while major sites with thousands of SKUs focus on refining their search engines to retrieve the very widget the consumer is looking for, improving the search function at Wilsons and other branded fashion sites involves more merchandising art than search engine technology. “As people went through our search engine, they were abandoning the site at very scary levels,” Hamilton said. Why? Because, the company concluded, it wasn’t putting its merchandising skills to work on the site the way it does in stores and brochures. So, apparel was organized by brand, not type, and product pages that once featured only the product being searched for now show that product displayed more appealingly in the context of matching accessories.
Back...
