Carrier Heats Up Web Sales Through Data Mining
By mining its customer database, Carrier analyzed its customers' buying habits and segmented them by ZIP codes. In May, Carrier's Web site, www.roomair.com, launched a new system that allows site visitors to input their ZIP codes and receive offers tailored to their needs.
Since then, the system has attained conversion rates of 7 percent, about 10 times the conversion rate the Web site had achieved previously. Online sales rates doubled.
To implement the system, Carrier hired WebMiner, New York, a data consulting service. WebMiner analyzed Carrier customer data and determined that the majority of its online customers lived in urban apartments or affluent suburban households. The analysis also found that a large segment of Carrier's customers were ages 30 to 37, Hispanic, used public transportation and lived in multi-dwelling buildings built before 1957.
WebMiner created this CRM tool by pairing Carrier's online customer database with demographic and real estate data. By using the ZIP code as an identifier, the company can guarantee the privacy of site visitors.
"We're only looking at neighborhood demographics," said Jesus Mena, CEO of WebMiner. "It gives you privacy and anonymity."
Previously, Carrier had relied on its server log files to count its site visitors and determine which keywords and search engines best drove traffic to its site. But the company found this type of analysis to be of little use in creating online sales.
Before implementing the data mining system, the Carrier Web site averaged about $1.40 in revenue per site visitor. In contrast, Carrier makes about $36 per targeted customer who receives a direct offer on the Web site.
Carrier's relationship with WebMiner began when Paul Berman, manager of global e-business for Carrier, attended a seminar given by Mena. Berman had read Mena's book on Web mining and wanted to learn new ways to grow the Carrier site.
"I was on a quest to find out how to leverage all the data we were getting on the Web site," Berman said. "There's a lot out there on data mining and Web technology, but nothing that really weds the two together."
Gathering information about customers can be difficult for Carrier, which relies on distributors to sell the bulk of its products to consumers, Berman said. The company has little direct contact with the people who buy its products.
Until recently, Carrier had access only to generalized market data not specific to its customers, Berman said. By selling its products direct and gathering in-depth data, the company is learning to build relationships with its customers.
"We're a 100-year-old company," he said. "It wasn't until the advent of the Internet that we were able to see our customer profile."
So far, Carrier has applied its CRM strategy only to its retail, window-mounted products, which represent a small part of the company's revenue. Carrier is working on ways to use what it has learned to improve sales of its installed residential and commercial products, such as central air-conditioning systems, that are available only through dealerships.
Carrier has about a 20 percent market share in this field, whereas its market share in retail products is only 1 percent to 2 percent, Berman estimated. Thus, Carrier sees its next big online opportunity in utilizing CRM and database strategies for business-to-business marketing to form relationships with the dealerships that sell Carrier products.