Evan Schuman's StorefrontBacktalk

The better—and far safer—approach is to let individual behaviors dictate app interactions. If any shopper starts using the interactive function repeatedly and uses it to connect with lots of contacts, by all means, react to that.

Profiling—treating any group differently based on their heritage (and even worse, a piece of software’s guess as to their likely heritage)—is never a good idea, especially when there is proof lying around.

Another is making sites more respectful and approachable to all groups. With Hispanic shoppers, though, that might not be so easy. A report from less than two years ago, for example, said that creating Spanish-language sites could easily backfire, if they’re not executed perfectly.

That conclusion involved two parallel concerns. First, a site that is not translated perfectly into conversational Spanish would end up offending more shoppers than it makes happy. But second, many Hispanic shoppers, according to that first report from Captiva, would prefer an American site that is in English.

Captiva CEO Lee Vann made that argument by breaking down the numbers, starting with the then-current figure of 30 million Hispanic consumers in the U.S. who are regularly shopping online.

“More than half of those 30 million actually prefer English. That takes the market from 30 million to 15 million,” Vann said. When he then removes bilingual consumers who are equally comfortable with both English and Spanish—a group that clearly neither needs nor particularly craves a Spanish version of a retailer’s site—that brings the Spanish-preferring number closer to 6 million. “And many of them are still comfortable with English,” even though it’s not their preference, Vann said.

Alas, no one ever said retail was especially easy.