PALO ALTO, Calif.--Baynote wants to eliminate the power of the blowhard on the Internet.
The Cupertino, Calif.-based company has come up with software that allows shopping sites or media outlets to better determine what their customers want by how people interact with their site.
If customers click on the specifications on a road bike offered on a sports site more than the specs of competing bikes, that extra bit of attention is an indication that the bike in question might be a bigger seller, Baynote says. The next time a customer comes to that site, that bike might be highlighted above other models, thereby increasing the odds that the bike will sell.
Tapping into this "invisible wisdom of crowds" through behavior is a far better indication of customer behavior than what people are saying in online reviews, focus groups, customer comments or comments from self-appointed influences on the Web, Jack Jia, Baynote CEO, said during a meeting at the AlwaysOn Stanford Summit taking place here this week at the university.
"There are three kinds of people that post comments--people with too much time on their hands, people who are too opinionated, and people with a secret agenda," he said. "Human behavior will tell you what (the general audience) wants."
Web site optimization is a crowded field, but Jia argues that the company's emphasis on unarticulated desires gives it an edge.
The company began selling its software about a year ago. Customers include Glam.com, eBay and Motorola.
One early customer, US Appliance, initially thought the software was buggy. Out of the large number of washers the online site sold, one of the most popular, according to Baynote's tracking data, was a red one. Weeks later, sales data indicated that it was one of the most popular models, he said. Red washers have also become a staple of the Best Buy Sunday insert since then, he added.
With media companies, the software is used for ad placement. "We know which ads are best suited for what spot," Jia said.
How customers react changes from Web site to Web site and product to product. Color might be a telling characteristic in appliances, but meaningless in sporting equipment. The company, however, does not emphasize clicks. Whether or not a person clicks on an item is a function of Web design rather than customer preference, he said. Instead, it examines about 12 different data points, which are determined by past consumer behavior on the site.
What are some of the common mistakes people make when designing Web sites? Sometimes sites are overloaded with content, which makes it difficult to find something, Jia said.
An inordinate amount of time is also often spent on user interface design. "They spend a lot of time on that rather than what people want," he said. "It can be pretty, but it can still be a pig."