Search engines and other e-retailers lose share as shoppers increasingly turn to Amazon for product searches, a Bloomreach survey finds.
Online merchants are refusing to accept as many as 16% of Internet orders in an effort to stop a tiny number of fraudulent transactions, experts say.
In an effort to block fraudulent e-commerce activity, e-retailers are blocking up to 16% of online orders--even though less then 2% of their transactions tend to be fraudulent, experts say. "This can cost them many more times the value of the fraud," says Jeff Foster, executive vice president of processing, consulting and risk management firm Retail Decisions USA.
Foster and other industry experts say that merchants in general block about 5% of all transactions due to suspicion of fraud, but that number rises to 16% for Internet transactions, particularly for transactions placed from foreign markets.
Foster says a common mistake that merchants make is setting ineffective rules for investigating or blocking transactions based on suspicions of fraud. For example, he says, a merchant might set a rule that all transactions valued at $200 or more be checked for fraud. But professional online criminals that monitor the activity on retail web sites often quickly learn that a site has set such a rule and take action to counter it. "If a merchant sets a rule at $200, fraudsters will place orders at $198 or $199," he says.
He notes that by blocking 16% of 100,000 transactions at, say, an average order size of $50, a retailer stands to lose $800,000 in sales in an effort to catch fraudulent transactions that may amount to only a small fraction of that. And they also risk losing the future business of legitimate customers whose transactions were unnecessarily blocked.
By contrast, he adds, a fraud-prevention software system that constantly checks for suspicious activity--such as multiple orders of products that are typically easy to resell, such as basketball shoes, jewelry or DVDs--coming from similar IP addresses, can cost about $17,000 per year.