Searching for Alternatives

It was a cold night in Pennsylvania when Leila Crooks was on Digg.com, the community-based popularity site, and came across a story about a "slanket" – a fleece blanket with sleeves that offers the freedom of arm movement so people can play video games or surf computers while snuggling under a blanket.

Crooks was intrigued, bought one, loved it and sent the link to three of her friends who all bought them. At $50 a pop, the maker of the Slanket was benefiting from Digg. The number of sites for finding information online – that are alternatives to search engines – is growing and the traffic to them is increasing. People go to them for different reasons: to find experts who can provide the best possible information, to have material presented in a different way, to see what other users value as important and to find information they know will be relevant to them specifically.

The Slanket example illustrates the difference between search (or "recovery") and discovery. Search (or recovery) is when you are looking for something specific – a confirmation of information that you know already exists, such as information about the governor of New Jersey or a recipe for meatloaf. Discovery is when you find something you were previously unaware of, weren’t specifically looking for or didn’t know that you’d have an interest in. It’s akin to reading additional stories in the newspaper because of the proximity to the article you wanted to read.

Amanda Watlington, founder of Searching for Profit, says social media sites like Del.icio.us or Reddit.com are organized to present information in different ways, which can appeal to people "depending on how their brains work." Users go to the Most Popular section of these sites and check out what others deem to be interesting.

Internet marketer Carsten Cumbrowski says he passes time on StumbleUpon.com, a browsing engine, to find sites that other online marketers find useful as well as to find sites that entertain him. He says it has a good filtering system – "if I say I don’t like something, I never get anything similar again."

Online marketers that want to leverage StumbleUpon can try its advertising system, which includes the link of the advertiser’s website in the regular StumbleUpon rotation. When a sponsored site is shown, a green button on the toolbar appears. However, some advertisers who have placed requests to get visitors in their category have received notices from StumbleUpon that there are not a sufficient number of people to view the ad in the category selected. Skeptics wonder if this is because StumbleUpon does not want to deal with a low ad spend or if they are overstating their traffic numbers.

Cutting Through the Clutter

As users become savvier in locating information, they realize that search engines are heavily monetized and loaded with nearly as many marketing messages as sought-after information, and they seek out alternatives, according to Sam Harrelson, general manager of the East Coast U.S., for Clicks2Customers.

Others agree that "less noise" and the struggle to find relevant information on search engines often lead people to alternative sites.

For example, if users are looking for tax help, they might go to Digg and read an article like "five ways to get your taxes done" rather than entering "tax help" into a search engine, which yields promotional sites about tax services, according to Chris Winfield, president of 10e20, an Internet marketing company.

Users often go to review or opinion sites to find information to complement what they have found on search engines. Winfield says he will search Google for a dentist in New York to get some names and then go to Yelp.com to look at their reviews. Searching for Profit’s Watlington says she searches for hotels in New York and then goes to TripAdvisor.com for the reviews.

Tim Mayer, vice president of product management of search at Yahoo, explains that when users don’t find the answers they want on search engines, they can ask a question on Answers.com. The site includes 4 million answers from publishers, original content created by its editorial team, community-contributed articles from Wikipedia and answers from WikiAnswers.com.

WikiAnswers is collaboratively written by volunteers "in the spirit of growing information for the public good," according to its website. For contributions that users find to be worthwhile, users vote with Trustpoints, which are indicators of how trusted the last contributor is as a member of the WikiAnswers community (as opposed to a measurement of how much you can trust the actual answer to a question). Trusting a user’s reputation is vital to not only WikiAnswers but to all social sites where users provide information or indicate the value of information (such as through tagging, bookmarking or ranking).

Just like in off-line world, the value put on information depends on who is giving it, and for this reason, users’ profiles can be weighty and influential. If you are reading an article about JavaScript on Del.icio.us, you look to see what other articles a user has saved – it gives you an understanding of that person’s knowledge base. It is similar to looking at someone’s book or record collection – it lends credibility and perspective.

Trust Me

Techmeme.com is one of online marketing expert Jim Kukral’s favorite sites because it decides what news is important as opposed to a site that simply aggregates feeds. "Techmeme saves me time. There is no need to go to a ton of blogs to figure out what is going on. That’s power to me," he says.

Techmeme works differently than other news sites. GoogleNews, a news aggregator site, uses its own software to determine what stories to display, but the sources are selected by a team of editors. Similarly, SFGate.com, the online version of the San Francisco Chronicle newspaper, also features stories decided on by editors. Techmeme creator Gabe Rivera explains that Techmeme uses a proprietary algorithm, which changes frequently, to analyze posts to determine what Web pages are being discussed or cited most often on the Web.

Blogger Robert Scoble (www.Scobleizer.com) explains that Rivera started by selecting 1,000 of the world’s top tech bloggers, put them in his server, studied their linking behavior and created a "fabric" that now includes thousands of blogs and websites. When Apple’s iPhone came out, high-profile bloggers in the fabric such as Michael Arrington (www.TechCrunch.com), Guy Kawasaki (http://blog.guykawasaki.com) and Scoble were all blogging about the new device. Because of this, the iPhone headline stayed up on Techmeme almost 24 hours a day over the summer. Scoble says he believes that information from a site like Techmeme is more valuable than information from Google because it’s more SEO-resistant – it is much more difficult for its links to be bought. For these top bloggers to link to each other, they must trust each other. "If I trust Arrington and he trusts Kawasaki and he trusts Joe Smith, then I am going to infer that I trust Joe Smith because my chain has trusted him. It would be very hard for a search engine optimizer to break into this chain," he says.

Dana Todd, president emeritus of SEMPO and SiteLab co-founder, says that she thinks it’s rather limited thinking to assume that all SEO is harmful and that SEO is the only market manipulation tactic on the Internet. "In any market, there are marketers – and they do exactly what marketers do. They attempt to find hype-holes in the system and exploit them." She notes that it took about 15 minutes for users of Digg to start manipulating the results. The findings of a September study by the Project for Excellence in Journalism (PEJ) warn that just because a news story is popular at a website (or within a certain community) does not mean that it is the most "important" story.

The PEJ study compared the headlines of user-driven news sites (including Digg and Reddit), and Yahoo News, which offers an editor-based news page and three lists of user-ranked news (most recommended, most viewed and most emailed), and compared these with the news agenda found in mainstream news outlets.

The study illustrates how the news looks different when audience members pick what story they want to read or recommend, as opposed to when a professional journalist makes the selection. The study found that the most popular stories on user-driven news are more fleeting and often draw on a controversial list of sources and reflect the interests of the participants in the community – stories on Digg and Del.icio.us tend to be more about technology, which is why they are popular among online marketers.

User-driven news isn’t new – in October the site Slashdot celebrated its 10-year anniversary as a site where users could scrutinize science, science fiction and technology- related news. It is credited for being one of the first sites to provide forum-style comments alongside user-submitted news stories. Just like Del.icio.us, you wouldn’t go to Slashdot to find out information on the latest U.S. billion-dollar defense policy bill.

In a post on his blog, Gaping Void, Web 2.0 writer/cartoonist Hugh MacLeod posits that if he were looking for a Vietnamese restaurant in Phoenix, he could Google "Vietnamese restaurant Phoenix" and possibly end up at a bad restaurant. Or as a blogger with a good-sized audience, he can ask about his dinner plans on Twitter or Facebook.com and get a couple of good recommendations within minutes. "Because I know these folks, or at least, they know me " there’s a certain amount of trust and bonhomie that comes with the recommendation," says MacLeod.

Social networking site Facebook has received lots of buzz and high financial valuation because of the "social graph" – a reference to graph theory that models the connections between things. Facebook founder Mark Zuckerberg says Facebook is not a social network but a tool that facilitates the information flow between users and their connections. It is the ability for users to get more out of their connections that people find compelling.

The Power of the People

According to Scoble, Facebook, Techmeme and Mahalo – a human-powered search engine that creates comprehensive and spam-free results for the most popular search terms – will kill Google in the next four years because users will get their information from these types of sites where trust is more than what algorithmic search results provide.

In October, Facebook added Facebook Flyers, which offers two different advertising options to the social network. Flyers Basic enables marketers to run ads on a $2 CPM with targeting based on age, gender and network. Flyers Pro lets marketers use pay per click with a minimum of 1 cent per click. As with other PPC ad buys, a higher max price per click increases the chance your ad will be shown. Online expert Kukral says Facebook Flyers "is basically Google AdWords within Facebook." He says that as an online marketer in Cleveland, this gives him the ability to do things like drill down to specific demographics with Facebook and target those users for a local "event."

"I can create an event in Facebook and then look at all the people on Facebook that are in the area and maybe have certain political or religious beliefs [based on their Facebook profiles] and then invite them to participate in an offer or event," Kukral says. "That is powerful and it could be the next big thing."

Part of trusting someone’s advice or being receptive to marketing messages is awareness of users’ tastes. David Rodnitzky, vice president of advertising at Mercantila – a collection of hundreds of online specialty stores selling to U.S. and Canadian consumers – says StumbleUpon and movie site Flixster.com are popular because they leverage collaborative filtering.

Here’s an example of how collaborative filtering works, according to Rodnitzky: Two users rate 200 movies on Flixster, and 90 percent of the time the ratings of the same movie are consistent (user A gives "Star Wars" a 10, user B gives it a 10). So if user A wants to see a Chinese-language movie but has never seen one before, and user B has seen five of them, the odds are good that whichever Chinese-language movie user B ranked a 10 will also be a movie that user A likes. When user B types in "best Chinese movies," the results are tailored to his specific likes and dislikes.

"Over time, if a collaborative filtering engine gains enough information about an individual user, it’s possible for the results to be very powerful – and far more accurate than what you get by just doing a search on a search engine," Rodnitzky says. However, the collaborative filtering engine first needs to have enough users to make those ratings viable.

Wikipedia.com is a popular search alternative that has garnered enough users to make it worthwhile. It reached 2 million answers in the English-language version in September 2007. Since starting in 2001, more than 100,000 registered users have made at least 10 edits each to Wikipedia articles. It is in the search toolbar in the Firefox browser and the sixth-most-visited network of websites worldwide. Internet marketer Cumbrowski claims Wikipedia is exceptional because it lists references – users can find out what experts think are the best resources for a topic – which obviates the need to research a topic any further.

Specialization

Finding the best information from the most informed user base is driving the growth in specialized communities and vertical search engines. Cumbrowski says there will be more specialization. As good examples of that, he points to BUMPZee.com for the affiliate community and Danny Sullivan’s Sphinn.com community for search information.

The explosion of content available on the Internet is fueling this specialization. Although Web 2.0 has made creating connections easier, it has made searching for information more difficult than ever. Because users’ queries are usually ambiguous, Google cannot serve the needs of every user. In turn, that has brought about an increase in the number of vertical properties, which restrict the scope of a search (see sidebar, Page 46).

Stephanie Agresta, a founder of the Conversation Group, says social networking sites such as Facebook and Twitter fill in the gaps. They allow individuals to tap in to different levels of networks of people to get information from someone who knows about a particular subject. Sites such as Mahalo and Squidoo.com enable users to view information through a specific user’s lens – the movement now seems to be toward a custom feed based on an individual’s friends and context, and away from algorithms.

But SEMPO’s Todd says it took only about 20 minutes for her to get bored with Facebook "because of all the ridiculous plug-ins and faux human interactions." She says that search engine optimization is not really the issue here. Google dominates that area because it caters to the very lowest common needs of users, and does so very elegantly. "It’s a tool, not a destination."

Moving forward, more of these types of sites are expected to pop up. Mixx, a new social news site – a cross between LinkedIn, Reddit and MyYahoo – is a social network that lets users find and share news based on their interests and location.

Another social network service is Ning, an online platform for creating social websites and social networks. Ning helps Web publishers create social networks around their content – more than 100,000 sites have used Ning’s tools to add their own networks. The sites range from a network of family members sharing content and photos to large networks such as Indiepublic, a social network for independent designers and artists.

Social sites are limited to certain topics, as several industries don’t have enough people using them yet and it’s tough to find any long-tail information on social sites, according to Web strategy consultant Cameron Olthuis. He expects that search engines and "alternative sites" will be completely necessary for people to continue to find information.