SearchReviews’ New Web and Mobile Tool Aggregates Millions of Consumer Reviews

2/15/11Follow @wroush

Every once in a while a startup pops up to lay claim to some area of search that Google, inexplicably, has not. (Yes, there are some.) This week it’s SearchReviews, a small Palo Alto, CA-based company that’s tapping the Web’s tens of millions of user-generated reviews of everything from electronic gadgets, to books to travel destinations. Today the company took the wraps off its new desktop search site as well as mobile search apps for iPhones and Android phones.

Founded by serial entrepreneur Ankesh Kumar, SearchReviews has collected more than 40 million reviews from more than 1,000 e-commerce and review sites, from Amazon to Zappos. Looking for an immersion blender that you can use to puree soups and smoothies? SearchReviews can locate 639 reviews (in this case, Home Shopping Network and Chefs Catalog are the richest sources). Searching for the perfect beach hotel for that Cancun vacation? SearchReviews has 1,178 reviews, mostly culled from TripAdvisor.

But more than just aggregating reviews into huge lists, SearchReviews re-indexes all of the reviews it collects to make it easier to answer specific questions, Kumar says. Say you’re wondering which Cancun hotels have the cleanest beaches. Turns out there are 94 reviews that touch specifically on this question.

It was just this sort of query, in fact, that led Kumar to build SearchReviews. “I was going on vacation for a few days with my wife and kids in Napa, and I was searching for hotels with indoor pools,” he says. “I wanted to find out how is the water temperature, how clean is the pool. I found five properties with 200 reviews per property. I don’t have time to read all of those.” Kumar calls SearchReviews a “mini-Google” that’s optimized for such keyword-based searches and saves people from having to search multiple review sites separately.

But it’s not stealing traffic from the sites whose reviews are indexed: the search results at SearchReviews lead users right back to the original source, such as TripAdvisor. The SearchReviews mobile apps add a dimension by allowing users to start a search by scanning a barcode with their smartphone’s camera. You could use this feature at Home Depot, for example, to gather consumer reviews before deciding which gas-fired outdoor grill to buy. There are also social features, such as a button that makes it easy for you to survey your Facebook friends for their own recommendations. “We have built it to be a Quora for shopping, if you like,” says Kumar, who adds that he bought his last minivan on the strength of a recommendation from friends who had children of the same age.

SearchReviews earns money, for the moment, through pay-per-click Google text ads on search results pages. As the company grows, Kumar says it plans to partner with review and e-commerce sites, which might bring in fees for lead generation and commissions on purchases. Kumar is funding the four-man startup so far out of his own pocket, using money he earned through previous ventures such as Sharetivity (a social search startup that sold key patents to Google), Personic (a job applicant tracking system bought by Kronos) and AT Systems (an IT consulting firm bought by Monster.com).

SearchReviews’ search results can be a little unpredictable for products outside the mainstream. For example, I scanned the barcode on the box for Tune Up, a program sold by a San Francisco startup of the same name that helps Apple iTunes clean up their track and album listings. SearchReviews got the name right, but it apparently couldn’t find any reviews of the software itself, so it returned an odd set of results that included Tune Up Cafe in Santa Fe, NM, and a Craftsman lawn mower tuneup kit.

But that’s probably just a sign of how new SearchReviews’ system really—Kumar says the whole idea was spawned just five months ago, in October 2010. And as the company moves beyond simply scraping the content of review sites to partnering with them, the data may improve. And Kumar’s ambitions are to make the site even more social, so that the algorithmic results would be supplemented by live advice from other people inside or outside the user’s social circle. “If you’re doing a search query about storage devices, we can connect you with someone else who did similar searches, depending on your privacy settings. At the end of the day, that’s what a review is—people on the Web being good Samaritans. That’s really where we want to get to.”

Here’s a video of Kumar demonstrating SearchReviews’ mobile app.

Wade Roush is a contributing editor at Xconomy. Follow @wroush

By posting a comment, you agree to our terms and conditions.