buying and reviewing, what they’re saying in their reviews, and what they do after reading the reviews, PowerReviews may be able to help e-retailers adjust their sites to help consumers locate popular products more easily. “The reviews are just the starting place,” says Chen. “What you see in the reviews, as a consumer, is probably 10 percent of the value.”
PowerReviews is Chen’s fourth company and his second startup. The first was FogDog, a San Jose-based online sporting goods retailer that started out in 1994 with the classic “four guys in a dorm room,” in Chen’s words, and went public by 1999. GSI Commerce bought FogDog in 2000, and Chen’s team moved to GSI’s Philadelphia headquarters, where they used FogDog’s technology to rebuild GSI’s e-mail marketing and shipping management back-ends. After GSI, Chen went to Yahoo Shopping, which he helped transform from a portal filled with mini-stores into a full-fledged comparison shopping engine that competed with the likes of Shopping.com and Pricegrabber.
Chen left Yahoo in 2005 with an idea for improving the whole online shopping experience. “Amazon had solved how to build and manage a great online store, but no one had solved how to use the Web as a shopping tool,” he says. “You still had to do tons of research and spend read magazines and spend countless hours figuring out the best products to buy.”
In some ways, Chen argues, the Web has actually made shopping harder. “Back in the days before the Internet, if I needed a new washer-dryer, I’d go to Sears, I’d ask a couple of questions, and they’d say you should chose between these two models. You’d take it home, it’d be great, and no one gets hurt. Now, you buy Consumer Reports, you read a whole bunch of reviews, you ask some friends, you do weeks and weeks of research, and if you buy it and it’s not the right one for you, you blame yourself.”
The “holy grail of e-commerce,” Chen says, would be a system that gives people lots of information but helps them make a confident decision faster. And the key to such a system, he thinks, is to structure the whole online shopping experience around customer reviews, rather than around classic search results. “You don’t go to Google to search for restaurants—you go to Yelp,” says Chen. “You don’t go to Google to look for a hotel; you go to TripAdvisor. The reviews on those sites break it down in a way that is far more compelling than any other option.”
Amazon certainly has lots and lots of customer reviews on offer, but it still organizes the products themselves as if the site were a collection of separate big-box stores: books, electronics, home & garden, toys, clothing, sports, et cetera. “Especially back in 2005, the point of view they were giving to the consumer was an Amazon-centric point of view,” says Chen. “Amazon still has not figured out how to use the users’ voices to actually guide people when they’re shopping. Reviews are still an add-on to the product pages, not the thing you use to start your process.”
The original business plan at PowerReviews was to build a “review engine” that would collect structured reviews from buyers after their purchase—inviting them not just to pen open-ended essays but to answer specific questions and name specific pros and cons for each product. (Today, some 95 percent of the company’s reviews arrive in response to the post-purchase e-mails that PowerReviews sends out on behalf of its customers. If you buy a tent from REI.com, you’ll get an e-mail a few weeks later asking what you thought.) The startup planned to offer the engine to thousands of specialty online stores that weren’t big enough to have their built-in review systems. With the permission of each retailer, it would mirror all of the reviews at its own recommendations site, Buzzillions.
PowerReviews built the review engine, and it quickly caught on. But the business plan didn’t work out, for a couple of reasons. “Because we had this competitor that started at the same time [BazaarVoice], we had to invest in a sales and marketing infrastructure to ‘sell’ the free product at an enterprise level, without enterprise-level revenue,” says Chen. At the same time, he says,
