Webware

SAN FRANCISCO--Google Vice President Marissa Mayer made a surprise announcement at the Web 2.0 Summit in San Francisco on Wednesday: "Social Search," a new Google Labs experiment that will bring in search results from a member's social-network contact circle.

It'll be launching as an opt-in project in the next few weeks. Then, you'll need to have a Google account and set up a Google Profile to fill in information about the social networks that you use. Google first launched Profiles about a year ago.

"What we've done here is inserted, on the bottom of the page, content written by people in your social network," Mayer said, adding that Google hopes this will "really improve the overall relevance, comprehensiveness, and quality" of search results. A search for a local restaurant, for example, could bring up your friends' Yelp reviews for the same establishment. A search for travel destinations could bring up a post from a friend's blog.

This comes on the same day that Google announced that it had entered into an agreement with Twitter to bring real-time "tweets" to search results. That's another product that has yet to actually launch.

"The idea is for...these fast-rising queries, where there's a period of time (when there are) actually tweets about that topic, and the definitive news source hasn't been written yet," Mayer said of the Twitter partnership, declining to disclose its financial terms.

This post was updated at 4:25 p.m. PT.

Nothing against Google or any other big search engine, but I think my friends are smarter than the rest of the world. When I want advice on a restaurant, a product I'm thinking of buying, or where to take my kid on a rainy Saturday, it's my circle of contacts I want that info from. That's social search, and I think it's got a big future. I've covered a few interesting products in that space, and today I'm looking at another one that's rolling out during SXSW: Aardvark.

Aardvark is social search meets instant messaging, which is a clever marriage. You send a query to Aardvark via your instant-messenger client. The system figures out which people in your network (friends and friends of friends) might be able to answer it for you, sends them messages, and then forwards you the replies.

Aardvark does a good job of find people to answer your questions.

(Credit: Screenshot by Rafe Needleman/CNET)

It archives everything on the Aardvark Web site, where you can also manage your friends and the topics you're interested in helping people with.

Aardvark's intelligence is the parsing and networking. It assigns categories to your natural-language queries and matches them to people who've indicated they can answer questions in them. I won't be getting fashion queries, for example, but I might get questions on places to take 2-year-olds in San Francisco. I also set it up so I only get questions when I'm online (per my IM status), so I don't get questions stacking up in my account.

In its early stage of development, it connects to AIM, GTalk, and Windows Live Messenger, but not to Yahoo IM. It also connects to Facebook. That's great--you don't have to start your Aardvark network from scratch. I found that 20 of my Facebook friends were already on Aardvark, and when I sent out my first query, I got replies back in minutes from people on that list as well as from friends of the person who invited me to Aardvark, co-founder and ex-Googler Nathan Stoll.

Who told Aardvark what I know?

(Credit: Screenshot by Rafe Needleman/CNET)

I was impressed by the speed and quality of the answers I got back to my sample queries. Also, Aardvark expanded on the three topic areas I put down that I was knowledgeable about with several more that were accurate. I think it got them from my Facebook profile.

You could of course use Twitter to send questions to your circle of friends, but Aardvark is better, since it sends queries to friends of friends, doesn't spam all your followers with questions they may not be able to answer, and collects and organizes the replies for you. Aardvark doesn't yet work with Twitter, but that's coming, as are Yahoo IM and SMS connections.

Stoll said Aardvark will open to the public "in a few months."

See also: Fluther (review), another IM Q&A service, and Delver (review), a social-proximity-based search engine.

On Thursday Yahoo rolled out deeper integration with Facebook as part of its SearchMonkey open semantic program. Now, when a person's public Facebook profile shows up in search results, there's a list of options on top that let you add them as a friend, send them a message, poke them, and see who they're friends with.

This (obviously) only works on Facebook profiles that have been made available for indexing by search engines, something Facebook introduced in late 2007. Users also have the option to opt out of getting their profiles indexed entirely.

One thing you can do to improve seeing someone's profile at the top of Yahoo results is add "on Facebook" to your search, as that's how Facebook formats the listing. In most cases, simply adding that brought up the person's Facebook listing to the top where it was otherwise showing up on the second or third result pages.

Instead of just a link to the person's profile, you now get quick links to various actions for Facebook. If you're signed into Facebook already, these options hop you right to the action.

(Credit: CNET Networks)

One of my favorite companies from the January, 2008 Demo conference was Delver, a search engine that takes into consideration who your friends are and what they've said and bookmarked in its results. I liked the idea in January (see Damn clever: Delver makes search social) and I like it still--I've brought it up in several posts since then.

I finally got a chance to try it out, as you can today, now that the site has gone into open alpha testing. At the current time, it's cooler in theory than in practice, but there is a ton of potential here.

Once you tell Delver who you are, it builds your social graph by itself. It correlates your identities across sites like MySpace, LinkedIn, Flickr, Facebook, Digg, and more, and creates a list of who your friends are. It also layers in a list of your friends of friends. Then, when you search for something, it gives precedence to content and links from your friends and their friends. So if you're looking for an Italian restaurant recommendation in New York, you'll get results from people you know, or people your friends vouch for. Delver CEO Liad Agmon clearly believes that online content from people you know is more valuable than generic Google results. He says, "The Web is no longer just a collection of documents. It's made up of microcontributions."

In this query, Delver gave me a search result from a guy who went to the same college as Josh, who's in the first ring of my social network.

You can also help Delver build your social search graph by feeding it sites and blogs where you hang out and have friends. Or to really supercharge it, feed it your FriendFeed page, where the "fusion" of multiple online personalities is already coded in. But you don't have to do this, and that's one of Delver's very cool features: It discovers your network on its own.

Delver can only extract results from people who post content or link to Web pages. If your friends are quiet online, you won't get much in the way of relevant personal responses. But you can, if you like, define anyone as your "search buddy," and then when you search, the results they would get become yours. You can even set up multiple search buddies to blend together the results from several people who have good online networks.

In using Delver, I liked how it told me the relationship of the person whose results it gave me. Some results came from my direct contacts, some from friends of friends, and some were selected because they were from people who went to the same school I did or worked in the same company.

However, in the alpha, I often got random (non-friend) results ahead of results from my social circle. Also, Delver doesn't index Twitter, and won't directly do so, according to Agmon. All those tiny posts would clog the Delver engine. Instead, eventually, Delver will clump Twitter posts into groups and index those intermediate pages. (Which doesn't explain how Summize manages to index Twitter.)

Coming later, possibly at the TechCrunch 50 event, will be a widget for bloggers: a "grey-label" search solution that gives blog readers an opportunity to get search results filtered by writers' social networks. Agmon says this will let bloggers "become prisms to the world," for their readers. We've seen custom search engines before (Eurekster, Rollyo), but this does sound like a nice add-on product for Delver.

Delver is well-funded, which is important since search is an expensive problem to solve. The company will make money the old-fashioned way: From search advertising. Agmon has not yet revealed which advertising network his company will use.

Related posts:
Google's view: Three trends in social networking
Social networking meets search: Sightix
All Webware stories mentioning Delver

In-browser social network Me.dium is expanding its services Wednesday night with the launch of a new social search tool. It pulls in regular old Yahoo results as part of the company's freshly announced BOSS platform (see news story here), while combining them with social results from other Me.dium users.

Me.dium founder David Mandell is calling the new system "Crowd Rank" and says it's not about how content links with other content, but how it links up with other users who are visiting these sites. Based on the data from people with the Me.dium sidebar or toolbar installed, the engine will get its own community-specific results that Mandell thinks will be more valuable than something merely indexed by machines.

That's not to say it's completely nixing those machined results. The social layer comes secondary to the service's main search, which will simply pull up Yahoo results. The extra value here is in the Me.dium community metadata that's wrapped around each link. Included is rank, velocity (how fast it's moved up in the results), crowd level, the last time a Me.dium user visited the site, along with how long most are spending there. It will be getting this data from two sources, both the social sidebar as well as a toolbar, which is launching as part of the service. Privacy will be the same for both products--as you can turn off tracking of sites you're visiting with two clicks.

Users looking for deeper integration with their in-browser search will have to use the toolbar or sidebar for the time being. Mandell says an option to use it in the top corner of compatible browsers like Firefox should be coming in the near future.

Related: Wikia Search launches the hackable search engine

Medium's social search will surf regular Yahoo results, while letting you search based on Me.dium user results too.

(Credit: CNET Networks)

I wouldn't say that Google is broken. But after looking at concepts by Delver (review) and Sightix, it has become clear to me how much better search can be--in theory--when it knows more about the person doing the searching, and when it knows about their social network.

Delver, a search engine that deciphers your social network, is still in private beta. Google is reportedly working on social search, but has yet to release it. According to Sightix, on May 1 it will deliver social search to Shin1, an Israeli social network.

Sightix shows you how your network influences your search results. (Click image for full version.)

(Credit: Sightix)

When you search for something using the Sightix technology on a social network, it searches the content of everyone you are connected with, and ranks results gleaned from each user based on the strength of their connection to you. Ari Gottesmann, vice president of Sightix, gave me this example: Say you're searching for a nightclub. The clubs your friends talk about will get a higher relevance than your friends of friends. It's much more likely that this ranking will yield results that work for you than searching a general engine that doesn't give extra weight to your friends' recommendations.

The product also works as a people search engine of a sort. If there's a good result from someone way out in your extended network, this tool will help you connect with them.

The Sightix company, originally in the business intelligence business with customers such as Dun and Bradstreet, has given its social search product a rich and complex interface. I think it's overkill for the task at hand, but if users want to see how they're connected to the results they get, it could be fun to explore.

It's less likely that this technology will work when users are looking for something completely obscure to their networks, but I bet they would be surprised to find how much they can glean from their extended social circle with a tool such as this.

Sightix is building its search product to embed in social networks. It's not making a search destination site, which means its success hinges on embedding the technology within social networks themselves. Getting the Shin1 deal is a good first deal, and Gottesmann is working on getting the service embedded in the huge global social networks for the future.

StumbleUpon, a fun Web discovery service, is getting a new feature, SearchReviews, that will take it beyond its historic role of hyperefficient time-waster. It could make it an integral part of the day-to-day browsing experience. SearchReviews pushes StumbleUpon ratings and other information into the search results pages on major search engines and content sites like Wikipedia, Google Reader, Flickr, and YouTube.

Users who run the StumbleUpon toolbar will now see little icons after search results on these sites that show them how popular a site is in the StumbleUpon community, the number of thumbs-up ratings it has, and who of their friends has rated the page.

The new SearchReviews features inserts StumbleUpon icons in your search engine.

It changes the way you look at search results pages. Now, instead of just looking at a sea of links in descending order based on what your search site thinks is most important, you also do a quick visual scan of the results for little StumbleUpon icons that indicate that actual humans have marked the page as worthwhile. If you happen across a site that a friend has flagged, you'll have an even better indication that the site is something that you'll want to check out.

If a page comes up in a results page that you've already flagged, you'll also see that rating (thumbs-up or -down) in the list of icons, reminding you of your own opinion of the page.

Other sites that get the SearchReviews feature: Wikipedia, Flickr, YouTube.

This is better than the social search that Hakia offers, for a few reasons. Most importantly, it doesn't require you to change your search engine. The new StumbleUpon icons show up on the search pages and sites you're already used to. And like Hakia, StumbleUpon will link you to discussion pages, but unlike Hakia, you get other social data such as the list of users who liked the page, so you can see what else they also liked.

This new feature makes the StumbleUpon social network much more valuable, so StumbleUpon is also making it easier to add friends from your other networks to it: You can import your friend lists from popular online e-mail services, as well as from Facebook. The service is also getting an Outlook contacts importer as part of this update.

Ironically, while SearchReviews is being integrated in 11 high-traffic sites, it's not going to show up on its parent site, eBay. I got two reasons for this from Garrett Camp, StumbleUpon's chief architect. First, he said, eBay items are "temporal," and StumbleUpon is more about bookmarking long-lived Web pages. Second, the 18-person StumbleUpon team couldn't handle customer support for eBay's giant user base if eBay were to actively promote the service.

For future releases, the team is looking at a completely general version of SearchReviews that could evaluate (and display ratings for) every Web link on a page.

Turn on these options to enable SearchReviews

StumbleUpon makes money by selling sponsored Stumbles. The new SearchReviews feature completely bypasses the ad engine of StumbleUpon, so it won't be a direct revenue generator. However, since it makes the entire service more valuable, it could drive more users to adopt StumbleUpon, and encourage current registered users to use it more. Personally, I uninstalled the StumbleUpon toolbar months ago, but I'm turning it back on so I can use this new feature.

See also: Mahalo Follow and Streakr (review).

Sproose is a relatively young search engine that now allows its users to control search results via voting. Each search result gets its own vote count and the option to click "I like it," which brings the result up to the top of the heap. The obvious comparison here is Digg.

Sproose searches through videos, using Blinkx, although these results cannot be re-ordered or voted on. Users can also opt-in to have their votes recorded on their profiles for others to see and vote on. All results get their own comment area, where users can talk--although I doubt many will want to write about a search result.

One of the more interesting inclusions in Sproose is its tag cloud, where users can browse for sites via tags. This is kind of an interesting form of Web discovery service, although it's lacking a clear way to add tags to sites yourself.

The obvious concern with sites like these is gaming results. Sproose deters some of this by removing votes from unregistered users after a limited time and requiring registration, although savvy spammers can usually find a way to circumvent whatever verification processes exist. Sproose's results are only as good as its users, and if security fails, so does the site.

See also: aftervote

Search results can be voted and commented on by users. The results with the highest vote counts go on top.

(Credit: CNET Networks)

Question-and-answer Web sites can connect the curious with the well informed, but some fail to live up to their promise. For instance, Yahoo Answers covers a vast array of subjects, but too often it attracts too many teenyboppers. With the new LinkedIn Answers, however, you can tap into the wisdom of a professional crowd.

This could be a great way of reaching out to those in the know without having to directly contact individuals who aren't so familiar with you. You can close a question to all but a small circle of chosen contacts. I also like that LinkedIn Answers displays topics posted and answered by people within my extended network. And lest hungry job-seekers clutter the pages, you're supposed to admit while posing a question if you're promoting your services, looking for work, or seeking to hire someone.

Many of the queries regard specific business matters, such as, "Which Source Code Management solution are you using and why?" Other questions are less concrete. Last time I checked, there were 68 responses to: "Will thoughts produce reality?" I left my question about the potential success or failure of Microsoft Office 2007 open to all of the 9 million members of the LinkedIn community, while alerting one software-savvy contact within my network. In less than an hour, I received a pair of thoughtful replies. Each query remains open for seven days before it is archived. I hope that LinkedIn Answers will attempt to eliminate duplicate questions as Answerbag does.

Overall, the uncomplicated interface is a snap to navigate. Still, while it's simple to look up contacts' names, I couldn't find an easy way to fish through the questions by subject. So far, however, along with its directory of services, LinkedIn's Answers appears to be a natural fit for this popular networking service.