In its second-largest acquisition, Amazon buys the company for $970 million.
Sharper Image’s back-end system that analyzes transactions for fraud red-flags 10% of orders across web and catalog. A team of customer service reps resolves 40%.
The Sharper Image has created 25 years of catalog experience combating fraud with offline analysis of transactions and it believes its system works better than online authorization systems, the company tells Internet Retailer.
Authorizing transactions offline helps keep the incidence of unresolved fraudulent transactions down to about 6% across web and catalog, The Sharper Image says. “We have a firm belief as a cataloger for 25 years that offline authorization on the backend is a much more effective way to control fraud,” says a spokeswoman. The Sharper Image batches all orders from its web site and uploads them hourly into its internally-developed fraud checking system. That system runs customer data on each order past checks such as dollar limits on purchase, address checks when billing and shipping addresses don’t match, and other limitations.
In addition to Sharper Image’s own data on each customer, the system uses outside sources such as the automated address verification systems used by most retailers. Sharper Image also keeps all data on any customers who have previously attempted fraud. Orders red-flagged during the authorization process--about 10% of orders--are suspended and routed to a dedicated team of a half-dozen customer service reps who call the customers for further information.
“We want to be careful not to offend anyone. The system does generate some false positives,” says the spokeswoman. About 40% of the suspended orders are resolved and those orders are processed, leaving the rate of attempted but unsuccessful fraud at about 6%. As to the actual rate of online fraud that slips through? "Because of our system, relative to other retailers we have a low fraud rate. The catalog business has really built up our system for checking fraud online,” says the spokeswoman.