|
Problem: Design Toscano wanted to increase response rates on e-mails.
Solution: Sent fewer e-mails, with higher product relevance based on customer transaction history.
Results: E-mail orders increased 800 percent.
As a marketing vehicle, Design Toscano's catalog does a fine job, generating about 80 percent of the home décor company's annual sales. Trouble is, the catalog is a general marketing vehicle: 100 pages of vastly different product categories that appeal to a broad cross-section of customers.
To bring certain products to the forefront, Design Toscano historically has used e-mails to announce catalog drops and highlight a handful of items, usually merchandise that's new to the book, says Kim Hansen, vice president of marketing at the Elk Grove Village, Ill.-based merchant. These e-mails also allow customers to preorder specific items that the company's merchandisers have deemed hot for the season.
Separate from catalog drops, Hansen sends e-mails to introduce merchandise lines exclusive to the Design Toscano site and announce site-wide sales. But while these e-mails effectively highlight new products and special deals, all customers with e-mail addresses receive the same message, regardless of whether they find these products or offers compelling. Hansen estimates that at one time, her customers received up to seven e-mails each month-and that was during nonholiday seasons.
Then in 2007, Hansen and her team began seriously testing segmented e-mail blasts. She already had some experience sending e-mails centered on broad product categories, such as religious statuary. A typical e-mail contained the headline "Express your spirituality!" and offered Christian, Buddhist and pagan statues.
"What we realized was that religious statuary wasn't as narrow a niche as we thought it was," Hansen says. "Religion, of course, is a personal thing, and just because a customer likes crosses doesn't mean she likes mystic fairies." She then split the single message into separate messages based on the style of statuary and built multiple lists from her customer database looking for individuals who purchased each type of iconography.
The results? Design Toscano's response on those e-mail campaigns was eight times greater than older, nonsegmented campaigns. Additionally, each customer received fewer e-mails, usually no more than four per month.
When Hansen first began segmenting e-mails based on product categories, pulling a list from the database was akin to pulling teeth. Up until that time, merchandise hadn't been coded in a consistent manner. Since, Hansen and her team have instituted a comprehensive reorganization of Design Toscano's transaction history. "In the past year, we've rigorously gone through the 15,000 items we've sold in the past 10 years and encoded the merchandise on a number of different dimensions," she says.
While functional categories, such as wall décor, statues and furniture, are maintained, additional layers have been added. Hansen now tracks the style, such as Egyptian or Gothic, and the genre, such as fantasy or religious. Hansen's next project is to tackle customers who've purchased across these various categories and assign them a score based on their likelihood to buy from an e-mail targeting one specific category or a versioned catalog mailing.
While segmentation at this level has its benefits, it's not without its challenges either. At press time, Design Toscano has one vendor that manages the customer database with transaction history and another that handles the analytics to produce the predictive models Hansen uses to create her versioned e-mails.
But as noted earlier, the increase in electronic sales is enough to drive Design Toscano to seek out new ways to give each customer a unique message.
Matt Griffin is a freelancer writer based in the Philadelphia area and former associate editor of Catalog Success.
|
