Shopping with a Wiki Touch

How do you follow an act like building an online advertising network, running it for 10 years and selling it to investors for $1.1 billion?

If you’re Kevin Ryan and Dwight Merriman, two of the founders of DoubleClick, you waste almost no time scoping out new targets of Internet opportunity. You do some market research, develop some proprietary technology and then launch ShopWiki.com, a new comparison shopping engine.

Wait—another shopping engine?

Yes, it’s a crowded field, full of players backed by big brand names like eBay and E.W. Scripps and plagued by low user loyalty to any one engine. But Ryan and Merriman, along with co-founder and CTO Eliot Horowitz, expect to be able to buck those odds, because they’ve come up with an engine with a difference—in fact, many differences.

“When we sold DoubleClick, Dwight Merriman and I went looking for another area on the Web that was big, but could be done better,” says Kevin Ryan, now CEO of ShopWiki. “That got us looking at Web shopping. If you asked me to come up with a better way of selling books online than Amazon’s, I’d say I can’t do that. But shopping was an area where we thought we could make big improvements.”

That conclusion derived from the advent over the last five or six years of major changes in both technology and online business models. The early shopping engines relied heavily on getting merchants to push their product and pricing information. Now, Ryan says, software can scour the Web for product pages and extract the relevant product information automatically. That extraction is not easy to do, but once that programming problem was solved, that too was a plus, since it threw up a barrier against competition from other engines that might also want to forego data feeds for a direct Web index.

For many of the veteran comparison engines, those data feeds also served as a prime revenue opportunity. If a seller didn’t pay, many of those engines wouldn’t display that vendor’s wares in the search results.

Half a decade later, not only has Google proven out the ad-revenue model of making money online, but it has conditioned users to expect a much more complete index of products in response to their search queries than any data-feed system could ever provide. “Imagine what search would be like if Google only included Web sites that paid a fee on its results pages,” Ryan says.

“We asked users what their fantasy online shopping engine looked like, and they told us they wanted the whole universe of products on there. That’s the point of shopping on the Internet; otherwise, people would just go to the store down the block.” At the same time, testers told Ryan and Merriman that they were tired of seeing screen after screen of merchant affiliate listings of the same product. So ShopWiki weeds out affiliates and lists only the parent merchants, showing the handful (ten or so) of the most clicked-on product listings first.

ShopWiki may still be in beta, but Ryan says it has already crawled the sites and extracted product data from some 120,000 Web merchants. Asked if any merchants have objected to having their sites scraped or their affiliates dissed, Ryan says not—and that since clicking on a ShopWiki product listing brings the user to the merchant Web site, most have been appreciative of the traffic boost. (A little snail icon next to some listings denotes a merchant page that’s particularly slow to load; but there seems more than a chance shoppers might misread that symbol to denote that a merchant fills orders slowly.)

Ryan says users also asked for an interface that would accept natural language queries with filters, so that a searcher could enter not just “diamond ring” but “2.5-carat square cut diamond ring”. A price slider at the head of the results page lets users adjust both the minimum and maximum prices of the item they’re looking for; those pages then refresh quickly and automatically to reflect the new criteria.

Then there’s that other tech innovation that wasn’t around when the first online shopping engines launched: the wiki. Wikis are Web sites or pages that let users add and change their content quickly. Wiki software has grown rapidly over the last few years as a means of collaborating online, and of course it’s the technology behind Wikipedia, the user-generated, user-edited encyclopedia favored by many.

Ryan and Merriman saw in wikis a way to help serve their users better. “We knew there were two kinds of shopping searches: one where I know what I want, and one where I don’t. I may know the make and model number of the specific digital camera I’m looking for. But I may also not know where to begin looking for the right camera. I mean, if I buy an air conditioner once every 10 years, how am I supposed to know which ones are best?”

So ShopWiki has made plans for a whole series of consumer-generated product guides on items their shoppers might want. Right now, the engine has primed the pump with about 1,100 guides written either by ShopWiki’s 20 employees or by knowledgeable freelancers. (Ryan himself penned the entry on what to look for in a ping-pong table.)

But the aim is to get the community of ShopWiki users involved in generating the product and buying guides. In fact, users can log onto a “community” section from the ShopWiki home page and try their hand at product guides that are still under construction or only in the suggestion phase. Unlike content on Wikipedia, those guides will always go through human editing, Ryan says, to make sure they’re not biased in favor of or against particular vendors.

Some of those guides are category-specific, such as what to look for in a high-end coffeemaker. But others are more thematic: “how to buy a present for a woman”, or “how to shop for Mother’s Day”.

So consider ShopWiki the consumer-facing shopping engine, built with the buyer in mind rather than the seller, Ryan says. “We’re designing everything from the consumer’s point of view. The pages are clean and the advertising—what few there are right now– are clearly identified as such.”

At the moment, those ads come from Google’s AdSense program, but Ryan says ShopWiki will consider signing onto other ad-serving networks too. “We’re really not focused on advertising right now. We just put those in so we could start learning about ads in this environment. Over time, there’ll be more ads, but they’ll always be done in a way that we feel users can be comfortable with.”

Of course, as a former head of DoubleClick, Ryan is comfortable with the notion of monetizing a Web site. But he says the company may never sell its own ads and certainly won’t do so in the near future—say, the next year.

“This is a heavily technology-oriented company,” he says. “Most of the employees are engineers, and for the moment we’re concentrating on building a great product that people are going to like.”

“Comparative shopping engines are naturally great advertising environments, and you don’t really need to worry about getting advertisers to the site. You need to worry more about having a service that people are going to want to use.”

But without data-feed capabilities, is there anything merchants can do to increase their chances of getting their product lines included in ShopWiki? “They can just remove any coding that prevents their sites from being searched and indexed by bots,” Ryan says. “Do that, and we be able to reach them.”