SES Chicago - December 7-11, 2009

July 27, 2007

Search Wikia Launches Open Source, Distributed Crawler

Wikia, the commercial site led by Wikipedia founder Jimmy Wales, is taking the next step toward launching its open source, human-assisted Web search tool, Search Wikia.

Speaking at the O'Reilly Open Source Convention (OSCON), Wales announced that Wikia has acquired Grub, the distributed search spidering technology previously owned by LookSmart, and will begin using it to build an index for the Search Wikia project.

Grub, which LookSmart bought in January 2003 for $1.4 million, is a distributed crawling service that LookSmart had implemented as a screensaver that would use idle CPU time on a user's PC to crawl the Web. The data was used to supplement its own centralized crawler's indexing efforts.

Wikia will immediately release Grub to the open source community, and make both the crawler and source code available at Grub.org. Users who download the application can run it either as a screensaver or a background process while other applications are running.

Specifics of the deal were not revealed, though it is part of a larger advertising deal between Wikia and LookSmart which was announced last week.

Under the deal, LookSmart will provide text and display ads in Wikia's freely hosted wiki communities, and eventually on the Search Wikia site, Wales said. Ads will be sold by Wikia on either a cost-per-click (CPC) or cost-per-thousand impressions (CPM) model. Inventory not sold by Wikia will be back-filled by ads from LookSmart's distributed ad network.

Share your thoughts on Search Wikia in the SEW Forums.

Posted by Kevin Newcomb at 12:30 PM | Permalink

May 31, 2007

Calacanis Launches Human-edited Search Engine

Jason Calacanis, former CEO of Weblogs Inc. and GM of Netscape, has revealed his latest project: a human-edited search engine/wiki/directory called Mahalo.

The idea is to hand-craft search results for the top search queries. At launch (in Alpha, meaning "not ready for users, but looking for feedback"), Mahalo has results for 4,000 terms, and expects to have 10,000 by the end of the year. There are currently 40 editors on staff, with 100 planned by the end of the year. Calacanis, currently an "entrepreneur in action" at Sequoia Capital, is the CEO of Mahalo. Investors include Sequoia, Elon Musk, and Newscorp.

Calacanis has no pretensions about replacing Google, since the human-edited model would not scale to match the billions of pages in Google's index. Instead, he looks at Mahalo as a way for searchers to get a better answer for top queries, while defaulting to Google for those not served by Mahalo.

"Google's mission is to index the world's information; our mission is to curate that wonderful index," Calacanis said in a statement. "It's my belief that humans can play a significant role in the development of search results and we're going to try to figure out exactly what that role is over the next couple of years. I am really looking forward to hearing what people think of the Alpha."

The idea is similar in certain ways to what About.com did with its guide sites, what the Open Directory Project was once, or what Squidoo or Wikipedia hope to become. Even with answers for only 10,000 terms, Mahalo's editors face a daunting task of keeping those terms updated, while adding newly popular terms.

But Calacanis has a track record of keeping a large team motivated, with more than 300 bloggers in the Weblogs Inc. Network at one point.

Another challenge, faced by all new search engines, is getting people to change their habits and try something new. Calacanis is counting on a good user experience and word-of-mouth to do that.

The site will be monetized with Google AdWords, and Calacanis has no immediate plans for Mahalo to sell its own ads. There is a way for users to suggest sites to a guide as answers to specific queries, but Calacanis – no fan of SEO – has said that process will be heavily monitored to prevent gaming of the system.

Track the coverage of Mahalo on Techmeme.

Posted by Kevin Newcomb at 10:06 AM | Permalink

May 9, 2007

User-Generated Videos On Your Domain?

Do you already produce or license video content for your site? Do you incorporate user-generated content and sharing tools now? Then encouraging your visitors to share video content on your domain seems like a natural progression.

Yet today we all equate video sharing with the portals. Sites like YouTube, Metacafe, Revver and Break have achieved critical mass, with YouTube far in the lead. They attract professional and amateur video contributors, who all host freely on these sites. Viewers know they will be able to search within these communities and find something that intrigues or amuses them.

While these video portals are doing well today, their predominance may not last forever.

A few days ago, YouTube announced that it would start paying some of its amateur contributors who brought in high traffic. While other sites had paid for non-professionals before, the market leader joined the fold too. (See earlier SEW posting.)

We're not cheerleading for particular services, but there are some low-end options for publishers. vSocial released a free service last year. CEO Mark Sigal says that nearly 600 users have created mini-YouTubes through his service so far. Vidiac provides another free option for small sites. Further up the food chain are services like VideoEgg, which powers some selected social networks.

It's worth trying to create more social engagement and loyalty on your site. At some point, the tide will turn.

Posted by at 1:30 AM | Permalink

February 15, 2007

ChaCha Ad Options Explored

ChaCha's "human powered" social search platform has been discussed here from the search side when it launched in September, and again in November when the site entered beta. Today, ClickZ media buying columnist Tessa Weggert takes a look at the ad opportunities ChaCha offers in "Introducing ChaCha Search."

ChaCha offers CPC-based text links targeted by guide-defined category, of which there are more than 100 available. ChaCha has also begun showing category-targeted display ads, including 300 x 250 rich media pre-roll video units, to help searchers pass the time while their guide does the searching.

Posted by Kevin Newcomb at 1:40 PM | Permalink

February 6, 2007

Eurekster Adds Community Features

Personalized/social search provider Eurekster has launched several new community-based features to its swicki platform. The new tools enable users to ask and answer questions within their communities, and those results will be added into search results conducted by other users through the swicki. More details at ClickZ.

A Eurekster swicki uses index data from Yahoo, Ask, Feedster, and other specialized search engines. It adds value with behavioral re-ranking and collaborative filtering, to create results that are more relevant to a given community. It previously did this through implicit factors, using click stream analysis to analyze user behavior, which is reflected in search results and the "buzz cloud," a display of recently searched keywords weighted by popularity.

The new features add explicit factors to the mix, engaging users to come in and contribute their expertise in a Yahoo Answers-type of way. Users can also write their own answer into the search results, which will be displayed once they are approved by the swicki owner.

Eurekster has also launched a swicki Ads service, which allows advertisers to buy ads on multiple swickis individually. Later this year, Eurekster will add a channel-based capability for advertisers to make category-based buys across similar groups of swickis.

Posted by Kevin Newcomb at 9:56 AM | Permalink

January 9, 2007

Toyota Gives Owners MySpace Like Blogs

The continuing growth of social media has influenced another market. Toyota has helped create a MySpace like program for their hybrid users to create a community of Toyota users.

This clever approach is one we are destined to see more of in the coming year. Not surprising - remember we are "social animals".

The increased importance on niched communities is finding support in many areas. Yahoo reports more about Toyota's move.

Gregg Benkendorfer, National Manager of Media Strategy and Digital Marketing for Toyota, told Yahoo, "the new community site taps into that excitement by letting users visually represent themselves through graphic art and video".

Posted by Frank Watson at 5:23 PM | Permalink

January 8, 2007

ChaCha Gets Big Boost in Funding in 2006

Lee Odden reported at Online Marketing Blog today that "Assisted-Search" provider ChaCha was funded with $6 Million in private funding organized by Bezos Expeditions in 2006. This story was also covered by Greg Sterling at Search Engine Land.

ChaCha has been the subject of an ongoing spirited discussion at the Search Engine Watch Forums. This has raised some interesting questions about the differences in search sophistication levels, and the availability of useful systems that can provide novice searchers with real-time human assistance not driven by marketing dollars and possible spyware incentives.

When ChaCha was first introduced to the forum, the initial response from Moderator David Wallace was something the majority of experienced searchers probably first felt when they learned of the system: why would someone want to wait while someone else searches for them when search engines can deliver results in nano seconds?

In fact, I also made a comment suggesting that maybe the product would be good for an older demographic, and that I would probably never use it. The idea that search engines could be better if someone was there to help find the answers may seem foreign to many that read online publications such as this one and the two mentioned in the introduction. We are all well beyond needing help searching, right?

Others have spoken of the disparity between expert and casual searchers. It exists within users of advanced Business Intelligence systems, as Google's BI expert Matt Glotzbach points out: BI systems contain much valuable information about a company's process, workflow, etc., but that information is untapped by the average employee. If you look at the traditional BI implementation, a very small number of users actually access the BI system and use its data, so the question becomes how to make this extremely valuable information accessible to the broader user base. (Source: Software Magazine )

Does this translate to Internet search? Many would think so. Perhaps familiarizing oneself with searching by using human help could lead to future ability to dig deeper and find the information that is behind the long tail keyword varieties that many advanced searchers perform within the first or second query.

Back to the SEW Forums thread, one of the next entrants into the discussion was a (somewhat self-painted) naïve evangelist who had recently begun performing the duties of being a guide for ChaCha, who claimed to get nervous while awaiting queries from unknown users of the system. This person provided a good glimpse at the supposed average profile of a ChaCha guide: someone who is interested in a topic and has gained some experience using the system in order to be able to help searchers find results for particular topics. It is a community in a way similar to DMOZ editorship, based on category familiarization, it would seem.

The rest of the thread is worth a read (although a warning that it is long and sometimes argumentative), as it delves into the overall usefulness of the search results provided by the guides. It also possibly helped to shape some of the ways that ChaCha could present itself in future forums, since the discussion got a little heated thanks to consistent pushing by one member in regards to its usefulness.

One thing that should probably worry ChaCha a little, however, is page 8 of the thread , which has been revived; partially it seems due to new affiliate arrangements for guides enticing them to receive compensation for Toolbar downloads. Understandably, this is a way to further monetize the system and help support it, but one Moderator makes some serious claims about the system's download and possible attachments.

Hopefully ChaCha will stay committed to providing a useful and safe community for inexperienced searchers to learn how to make queries. Additionally, it would be nice if they disclosed clearly that some/many of the results presented are sponsored. Lastly, please get rid of any potentially malicious spyware in the download packet if that allegation is true!

(added 1/9: ChaChaInsider at SEW Forums denies that any sort of Spy/Malware is attached to the toolbar download - great news.) Hopefully anyone wishing to comment on this will do so in the referenced forum discussion.

Posted by Chris Boggs at 4:29 PM | Permalink

November 30, 2006

How The Digg Editorial Process Differs From Search Engine Editorial Process

Todd Malicoat has a write up called The Search Marketer's Guide to Digg, where he explains the difference between the Digg.com editorial process and search engines editorial process. It all comes down to the "human editorial authority," and I quote;

While most the search engines DO have human intervention - they haven't accepted and embraced it. One of the beauties of digg is if there is CRAP in the index - you know exactly who to blame for it.

As always, both human and algorithmic methods of intervention have their faults. I am sure Danny will go into a bigger write up on the pros and cons of each at a later point.

Postscript From Danny: I may write-up more on this in the future, and I talked about it on a recent Daily SearchCast episode. The short story is that it's been amazing to watch Digg effectively go through the same type of spam fighting evolution that the search engines have done.

Consider from Digg:

Digg is a user driven social content website. Ok, so what the heck does that mean? Well, everything on digg is submitted by the digg user community (that would be you). After you submit content, other digg users read your submission and digg what they like best. If your story rocks and receives enough diggs, it is promoted to the front page for the millions of digg visitors to see.

The overall idea is that the community does everything. In reality, there is a lot of backend editing and changes done by moderators. That's because the community, if left to itself, will have a small number of people who try to manipulate Digg for their own benefit.

It's an old story. Consider from Google:

PageRank performs an objective measurement of the importance of web pages by solving an equation of more than 500 million variables and 2 billion terms. Instead of counting direct links, PageRank interprets a link from Page A to Page B as a vote for Page B by Page A. PageRank then assesses a page's importance by the number of votes it receives.

Except it's far more complicated than that. Links have to be weighted, not trusted and entire sites removed because of spam and manipulation.

Both Digg and Google (and the search engines before it) started out in what I'd call "trusted mode," where you are optimistic that a community (people submitting; a collection of pages) can be trusted. Along the way, they shift to "mistrust mode" where you realize you need to be initially dubious about everything that flows in.

If I had more time, I'd go through and do a long compare-and-contrast on how recent Digg changes have exact counterparts in the crawler-based search engine world. Honestly, there are times when I could do a search and replace for the word Google to the word Digg in an article on spam fighting and the description would be the same.

The answer, by the way, is simple. Machines that the search engines depend on are imperfect (in particular, rankings can be manipulated more broadly), as is the human model Digg uses (in particular, humans can miss a lot of things). The combination of the two is much stronger. Some more thoughts on this from me:

And here are a bunch of related stories from across the web that we've included in our headlines recently:

Posted by Barry Schwartz at 9:19 AM | Permalink

November 6, 2006

ChaCha Search: Take Two

When "human-powered" search engine ChaCha launched in September it was a great idea with very flawed execution. The site offers algorithmic search, but also IM-chat based search with actual people in real time. Here's my previous post on ChaCha. Today, the site has gone from "alpha" to "beta" based on lessons learned during the alpha period. And it appears from casual searching there have been some significant improvements.

According to the press release issued today, "Since launching the experimental Alpha version in September 2006, more than 14,000 people have entered the system to serve as ChaCha guides with as many as 1,500 guides being added on any given day. The fast-growing guide community has been instrumental in ChaCha's accelerated Beta development, as they have been working around the clock to help end users find information online."

I spoke to CEO Scott Jones late last week about the discoveries and improvements ChaCha had made during the past few months. He said he was surprised by the number of people using guide-based search. "I would have thought 95% would search without the guide, but it's been more like 40% using the guides."

Jones said that ChaCha's guides are drawn from three primary pools: retirees, college students and stay/work-at-home moms. "I thought our target market for guides was going to be college students, but the thing that took off was work-at-home moms."

Guides are qualified and paid according to an elaborate system. But Jones explained that guides have the capacity to make as much as $20,000 per year at the top end. But he also said that they've instituted strict controls to ensure quality and are working to get better and better guides into the system.

The guide-discovered search results are integrated into the algorithmic results and given priority in subsequent searches. I conducted four searches to test ChaCha in beta:

These were all traditional searches without using the guides and the results were quite competitive I thought. The site was essentially not working in Firefox, but worked fine in IE. In a couple of cases I tried these same searches using guides but they were busy; however algorithmic results were provided. In one case I did have an interaction with a guide.

Here's the transcript in the context of a search for "Best non-toxic household cleaning products":

Status: Looking for a guide ... Status: Connected to guide: melindam melindam: Welcome to ChaCha! You: Best non-toxic household cleaning products melindam: Hi, how are you? You: fine melindam: I'm sorry, it says you're searching for Thanksgiving recipes. Let me change that and I will look for you. You: that was my previous search You: sorry melindam: No problem :-) You: now I'm looking for something different melindam: Ok, one moment please. melindam: I appreciate your patience while I find exactly what you need. melindam: Look at that one while I find more please. (Result found: mrsmeyers.com) You: That's a commercial result You: Mrs. Meyer's is a brand not an "objective" source melindam: Ok. Are you wanting a personal opinion? You: I'd like a reputable source that gives me a range of products and evaluates them, thanks. melindam: You asked for non-toxic cleaning products. Are you looking for reviewes or products? You: The word "best" implies an evaluation melindam: Well the word best is subjective. melindam: All companies say they have the best. You: Now we're having a philosophical argument You: "Best" can be subjective You: but it can also be a judgment resulting from tests You: or expert evaluation melindam: Let me transfer you to someone who may be give you better results. Transfer: You are being transfered to another guide who can help you search even better! Looking for guide ...

I didn't pursue it with the second guide. While I was waiting and interacting with "melindam," there was video running in the upper right of the screen. This offers a brand advertising opportunity and takes some of the pain out of waiting for the human to conduct the search.

The guides represent both a burgeoning social network with its own possibilities and a word-of-mouth marketing force. These are two elements that suggest the site will develop some staying power and find a market.

Jones and I talked for awhile about brands and search and the role that brand plays in search loyalty. He told me that he spend considerable time thinking and working on the ChaCha brand, "Cha means search in Chinese, and it's a dance -- to achieve a search result."

What are the most popular searches on ChaCha?

  1. google.com
  2. amazon.com
  3. digg.com
  4. weather.com
  5. ask.com

Posted by Greg Sterling at 9:19 AM | Permalink

October 25, 2006

Next Up For Sale, Digg.com

Techcrunch reports that Digg.com is looking to sell the property for $150 million. Digg's recent discussions to sell included negotiations with News Corp, and some other non-disclosed bidders. The word is that Digg wants "at least $150 million" but no one is willing to pay that much to them.

Specifically one of the reasons mentioned is because Comscore figures claim Digg has "1.3 million monthly unique visitors and flat growth since April" but Digg claims they have "20 million unique monthly visitors and steep monthly growth." Danny reviews Rand's SEOMoz article named Website Analytics vs. Competitive Intelligence Metrics that concludes "no publicly available competitive analysis tool we're aware of provides solid value," including Hitwise or other reporting engines.

So does Digg have 20 million visitors? I do not know. It is hard to measure without direct access to one's server logs and even then things can be spoofed. Advertisers and those looking to sell their sites always want to represent the highest numbers as possible here. But can you trust any of them? There are just too many variables and ways to fudge those figures.

Postscript: Matt Tatham from Hitwise sent me some recent data on Digg.com, so I thought I would add it here.

Recent US data on Digg.com in regards to their potential acquisition talks: -Digg.com's US market share of visits increased 231% comparing the week ending October 21, 2006 versus the week ending October 22, 2005 -Digg.com US market share of visits increased 176% comparing September 2006 versus September 2005 -Digg.com is the third most visited website within the Hitwise US News and Media - IT category for the week ending October 21, 2006 -Digg.com is the ranked at 114 most visited website within the Hitwise US News and Media category for the week ending October 21, 2006 -Digg.com received 55% of its US traffic from Google for the week ending October 21, 2006

Thanks Matt!

Posted by Barry Schwartz at 8:47 AM | Permalink

October 11, 2006

Url.com - Results Ranked By Users

The ability to rank or rate results delivered by search engines is slowly becoming more commonplace, and I'm noticing that one or two multi/meta search engines are either incorporating this functionality into search or indeed making a specific feature of it, such as the new URL.com.

I wrote about Jatalla on my own weblog at the beginning of September and was less than complimentary about it, since it had indexed very few pages (it has to be the only internet search engine that doesn't return any results for the search 'porn'; I was getting desperate to see some results, indeed any results by that point!)

However, since then we have seen the arrival of URL.com, which is an interesting and memorable domain name, though not necessarily one that I'd associate with a search engine.

Their tag line is 'search with many' and relates less to the search engines used and more to the users. Basically it's a search engine that pulls results from Google, Yahoo and MSN, ranks the results on the screen according to position and allows users to rank or comment on what they see. On the whole it does the job reasonably well, and is worth taking a look at.

Users will get the most out of the search engine if they spend the few seconds it takes to get an account, though this does mean that others will be able to see what your interests are, though this can be easily overcome by choosing a John Doe account name.

Searches ran reasonably quickly, although I noticed once or twice that one or other of the three was slow to respond, leading to a less than accurate set of results, though of course this is a problem inherent to this type of search engine. Rather more worrying however was the fact that I noticed that sometimes a page would be returned as being in fifth position in the Yahoo results for example, when it was actually first. This really is a fundamental flaw and should be addressed quickly.

However, users can click on the result that they want to view and the page is pulled into a URL.com frame, at the bottom of which the searcher can vote for the result ('good result' or 'not so good' and can also comment on the page. It's then possible to either close the frame and go directly to the website page in question, or back to the results page. If other users run the same search they will be able to see that particular individuals have commented on and/or liked/disliked the result. If other people comment on the page these are emailed to other commentators as well, hence the 'search with many' aspect of the site.

This function is easier displayed than described, so try a search for 'search engine watch'. Pages with positive votes will show up more than pages with negative votes, with the idea being the URL.com community will police what appears on their screens.

This is useful and interesting, but of course it's also open to fraudulent use. While this doesn't appear to have happened with the results yet I'm fairly confident that it won't be long before people with grudges against certain sites, or who want to obtain a commercial benefit from boosting their own sites get involved. One could always argue that the power of the many will overcome the comments of the few (that almost sounds like a quote from Mr Spock), but I'm not entirely convinced that will necessarily be the case. Interestingly there's nothing in the site documentation that I could find that addresses this issue, but I can see so many legal issues being raised with this it's painful.

All said and done however, it's a nice optomistic idea that works well - at the moment. If nothing else, it's a quick and easily memorable URL to allow searchers to get access to 3 of the major search engines.

Posted by Phil Bradley at 9:47 AM | Permalink

October 10, 2006

Court Issues Notice To Google For Allegedly Spreading Hate In India Via Orkut

The Times of India reports that Google is in hot water over Orkut once again, this time for allegedly spreading hatred for India. The article says, "The Aurangabad bench of Bombay High Court has directed the Maharashtra government to issue notice to Google for the alleged spread of hatred about India by its social network service Orkut." The problem is over a "We Hate India" community in Orkut that posted a picture a burning Indian flag.

Posted by Barry Schwartz at 8:35 AM | Permalink

September 19, 2006

Why Is Wikipedia Advertising On Google?

Wikipedia Advertising Shows Up on Google from Micro Persuations covers how Wikipedia is apparently advertising content on Google. A search on crowdsourcing in Google brings up a Wikipedia ad in the sponsored listing area that says:

Crowdsourced Software 1. Harness this exciting wave. 2. Save money or make money! wikipedia.org/Crowdsourcing

Yes, the same landing page is shown on the organic listings as well. So why would the Wikipedia be spending money on those ads? Is it even Wikipedia who is purchasing those ads? Maybe it is a 3rd party that is interested in getting more content added to certain Wikipedia pages. Maybe there is some financial interest on that page for some company?

Postscript Barry: ClickZ has a quote from Wayne Saewyc, a Wikimedia spokesperson saying, "As far as I am aware, the Wikimedia Foundation is not purchasing Google ads, or any other advertisements for that matter."

Posted by Barry Schwartz at 9:08 AM | Permalink

September 14, 2006

Checking Back Links From Wikipedia Articles

Steve Rubel spotted a new page at Wikipedia that allows you to quickly locate all the links pointing to a specific site from within the Wikipedia. This page allows you to enter in a domain, and it will bring back all the pages that have links pointing to that domain from within the Wikipedia. For example, Search Engine Watch has 107 citations from the Wikipedia, just search using the syntax *.searchenginewatch.com to find out. It is a great way to see what types of Wiki links you have and also to see how people are using your content in a positive or negative way towards your brand.

Posted by Barry Schwartz at 8:46 AM | Permalink

September 12, 2006

Big Brands: Do You Know What Wikipedia Saying About You Via Google?

Steve Rubel produced a small study on the top 100 advertisers, according to AdAge, to see where in the Google rankings does the brand's Wikipedia entry fall. For example, a search on the well-known automobile brand Chevrolet shows a Wikipedia entry for them at the number four result. What that Wikipedia entry says about your company can have a huge impact on your brand, given that Google will be driving traffic to Wikipedia. The study showed that the Wikipedia entry listed in the Google results for the top 100 brands, on average was at position 11. But Steve Rubel explains that many of those brands have listings within the top 10 and some even the top 5 for their brand name.

Steve explains that if you have a lot of domain names for your company, then the Wikipedia entry will most likely be pushed down the results. Also, he said Yahoo, MSN and Ask all have similar responses, but they are not documented in his study. He finally links and summarized an adAge.com article named Your Brand on Wikipedia.

FYI - this is part of online reputation management.

Posted by Barry Schwartz at 8:41 AM | Permalink

September 7, 2006

Digg To Change System To Stem Gaming; Top Digger Quits

Kevin Rose, founder of Digg.com, wrote that the Digg.com algorithm for ranking top stories on the home page and other category pages will be changing. Why? Because of other bloggers beginning to publicize that digg is rigged or at least favors the top 30 or so users (diggers). This notice of an algorithmic change to come to Digg.com has encouraged the top user, p9s50W5k4GUD2c6 to leave Digg with a comment he left here.

To be honest, I feel that this may be a good thing. Changing up the algorithm is a good idea. I know many people in the search community that know the tricks to 'gaming' the Digg system. Kevin Ross, in my opinion, is right in making a change to the system. Of course, this will upset the top Diggers, heck, if my name was removed from one of the fastest growing sites on the web, I may be upset too.

Seriously, read this to get an idea on how the top stories currently get ranked at Digg. Interesting, no?

So now Andy Beal reports that our search community has started their own niche version of Digg at Search N Sniff, will it take on - I don't know.

Digg.com is not the only social Web 2.0 community going through growing pains. Ben at the Search Engine Roundtable, reported that Facebook.com Users Riot Over New Features. Basically, new features that detail the "tracking of friends in users networks" has gone a bit too far in terms of privacy concerns and users are very upset about.

Social networking sites are feeling the pain. User generated content is generated by volunteers. If you upset those volunteers, they can get up and go somewhere else. Will they last?

Posted by Barry Schwartz at 8:50 AM | Permalink

September 5, 2006

Netscape Search Inserts Netscape News Above Web Results

What's this? Netscape Search has changed? So says Netscape's Jason Calacanis. What's new? From what Jason says and I can see, the big difference is that there's a new "Netscape.com results" section at the top of the page that shows you top voted stories from the Netscape community news service.

More on that service is cover in my Netscape Aims To Be Digg 2.0, Slashdot 3.0 With Community News Mode post. After the Netscape.com results, you'll get Google-powered results from across the web.

To be clear, this isn't a case of users voting particular stories to the top of search results, say on a query-by-query basis. Rather, from what I can tell, stories get popular over at Netscape.com itself. Then the top stories over there might show up in response to searches at Netscape Search, if they are a good keyword match. In other words, think of it as NetscapeRank. If a story has good NetscapeRank, that might help it rank well in keyword search results.

FYI marketers, it might not take much to pick up that rank. A query on cars gives me three results that have only one or two votes each. More than votes might be involved, of course -- and I haven't seriously poked at this in any details. But there may be opportunities, though the Netscape traffic is far, far less than at the major search engines.

From a searcher perspective, I guess it's mainly useful if you want to be alerted to news results in the course of your regular web searching. Places like Google and Yahoo have long done this already. The difference with Netscape is that you can get your news sourced from a community, rather than an algorithmic search engine.

Then again, use the Google Co-op link for Digg you'll find here, and then you'll sort of get what Netscape is doing, just with Digg as your news source showing up at the top of Google searches.

Perhaps down the line, we'll see Digg follow Netscape's lead and make it possible to search Digg and the web at the same time, something that doesn't seem possible right now at the Digg site.

Finally, search.netscape.com is a long-standing address for Netscape Search. That's still active and redirecting here, where the "old" Netscape search remains active.

Posted by Danny Sullivan at 9:18 AM | Permalink

September 4, 2006

New Engine 'ChaCha' Offers Real-Time Answers From Live 'Guides'

Part Wikipedia, part Yahoo Answers and part About.com, ChaCha is a new search engine with a compelling hook – real-time results from human beings. The site launches (in "Alpha") today and offers users two ways to search: traditional algorithmic results or help from live "guides." Users interact with guides via an embedded instant messaging window in the search results page.

Brad Bostic, co-founder of ChaCha, said that the site had lined up about 2500 guides at launch: college students, retirees, stay-at-home moms and others "who are online all day anyway." But not just anyone can become a guide apparently; you have to be "sponsored" (invited by an existing guide) and work your way up a hierarchy consisting of four levels.

New guides are considered "apprentices" and are matched with areas of personal interest and expertise. New guides also have mentors, more experienced guides who monitor their work. Apprentices cannot interact with the public initially and must pass several tests for speed, quality and accuracy. If they meet these requirements they become "pros."

Pros then get the chance to interact with the public and will be paid (US$5 per "search hour"). After pro come two other levels: "master" and ultimately "elite." Elite-level guides make US$10 per search hour. But once you become a master you're eligible to earn 10% of what your "network" makes. Your network consists of those you've brought into the "ChaCha Underground" (the community of guides).

The challenges of this entire concept obviously revolve around the cost structure and how many guides ChaCha can recruit to make the real-time aspect of this work well. It was clear from my conversation with Bostic, however, that ChaCha has carefully thought through these issues. The company has developed financial and ego-based incentives to recruit and retain guides and various mechanisms to help maximize the quality of their results.

While Bostic believes that most people will get and stay involved because of altruistic reasons (think Wikipedia), the modest financial rewards and four levels of "initiation" (my term) may add additional appeal for prospective guides. As mentioned, there's also a community aspect. Guides have profiles, featuring their areas of expertise and most recent answers. They're also rated by users.

The site is ad supported and already has a number of display advertisers as well as sponsored links. The key will be to keep the ad revenues climbing higher than the contractor fees. But Bostic assured me that ChaCha had crunched theses numbers many times. (This ChaCha Underground, if it grows, becomes another social network and advertising vehicle in itself.)

Like most new site launches of late, the first few days are likely to be shaky. I performed a number of searches with very mixed results. I had trouble getting algorithmic results several times and was only able to connect with a guide once. This was disappointing but probably due to last-minute engineering issues and fine tuning before launch.

The following is a verbatim IM transcript from my single interaction with a guide (a male who's name I've replaced with "guide"). My query was "Best LA hotel to stay in with kids?"

Guide: Welcome to ChaCha! Please wait a moment while I search for your results. Guide: Please hold a moment. Guide: I will find a good result for you. You: still looking? Guide: I appreciate your patience while I find exactly what you need. Guide: I am looking for details on kid-friendly hotels. You: thanks Guide: I have found several but will soon have one that is well-suited for your search. Guide: Do you want 5, 4, or 3 stars? Guide: hotel rating that is. You: how about most stars for under $200 per night Guide: OK - one moment. Guide: how many beds? You: 2 Guide: ok. Guide: Kids stay free at these. You: okay, thanks Guide: Let me check on the rates. Guide: The nice thing is that these both have full suites. Guide: So if you are with kids, you have refridge, etc... Guide: Is that good for you?

The entire interaction and elapsed time of the search was a little over three minutes. (Users will also automatically be offered algorithmic results while the guide is searching.) The results he ultimately sent were two hotels from an aggregator site. In fairness this is a difficult query to answer effectively in a short time frame. The guide doesn't know my personal preferences and he didn't ask where in Los Angeles I wanted to stay. But I did get two reasonable results.

The promise of people answering search queries will likely bring out qualitative and opinion-based questions, which is also what's going on over at Yahoo Answers. Fewer fact-based queries will likely be directed at guides because they are easier for people to answer on their own through conventional search.

Unfortunately I wasn't able to connect more than once to check the performance of the guides on other types of queries. But on Friday when I was briefed by Bostic, he and I went through two successful sessions with guides in real time. One was fact based and one opinion oriented.

Social search engines, which have been proliferating, are trying to leverage human expertise to offer a more relevant index or qualitatively better search experience. Chris Sherman recently wrote "Who's Who in Social Search," which provides an overview of the different categories and different companies in the space. In many ways, however, ChaCha's concept of guides providing real-time answers is the ultimate expression of social search.

One of the challenges that MSN Search has faced since it launched is that it's not obviously different or better than Google or Yahoo and thus it has struggled to gain market share. My belief is that people have search habits that are well established at this point and getting them to change is harder than most people believe.

But ChaCha does have an obvious differentiator in the form of its live guides. Many people will immediately like this concept and the experience of interacting with a person in conducting online research. But as Bostic said to me, "The Devil is in the details." Indeed, the actual performance of the guides and the quality of the ChaCha search experience as a whole will determine adoption or its opposite.

Because of the network of human guides, ChaCha is not easily or quickly duplicated by others, including the major engines. But they will be watching to see how ChaCha performs and how people respond.

ChaCha will not rise up and supplant market leaders any time soon. Yet if the site can recruit enough guides and make the live search experience fast and effective it will gain adoption. But the ultimate test will probably be determined not by any rational comparison of search results but by something more intangible: whether users think that ChaCha's live guides will do better than they can by themselves.

Post Script: Research master, Ask director and librarian extraordinaire Gary Price previously wrote a long post about Q&A services here and real-time library based search assistance here.

Posted by Greg Sterling at 4:45 AM | Permalink

August 31, 2006

Understanding Digg.com

Valleywag has an excellent Cheatsheet on What is Digg? Basically, if someone asks you what is Digg, just send them there. It describes the basics, how it works, how to break it, Digg's competitors, the recent publicity on Digg, "How Digg fits the buzzwords," oh and don't say "I got digged," I believe it is "I was dugg." Full details on Digg at Valleywag.

Posted by Barry Schwartz at 9:24 AM | Permalink

August 15, 2006

103 Links About SES San Jose 2006 (AKA The Big Recap)

Couldn't make it to last week's monster Search Engine Strategies show in San Jose? Well, maybe next time! In the meantime, I've compiled a list of coverage from across the web, even somewhat organized into topic areas.

Our San Jose show is always tough for me, as I arrive a week earlier to visit with the various major search engines out there. That means two weeks of news and email to dig out from, since you can never get it all done on the road. All that digging out means I know I don't have everything listed below. But you'll find plenty to keep you entertained.

General Recaps

Eric Schmidt Appearance

Eric Schmidt & Search Privacy

Click Fraud Panel & Related Coverage

Yahoo's Panama Ad Platform Preview

Social Search & Related Topics

Organic Listings Sessions

Search Advertising Sessions

Issues Sessions

News, Blogs & Public Relations

Big Sites/Budget Sessions

Small Sites/Budget Sessions

Conversion & Metrics

Other Sessions

Google Dance & Parties & Pictures

Posted by Danny Sullivan at 4:50 PM | Permalink

August 7, 2006

Yahoo Launches Search Builder

During the Social Search Overview session, Tim Mayer of Yahoo announced the launch of Yahoo Search Builder. I haven't had much time to play with it yet, and probably won't until next week. But the Yahoo Search blog has a nice overview of the new product. Keep in mind, this seems very similar to Eurekster, based on my quick quick quick read of it. More on this later.

Posted by Barry Schwartz at 1:02 PM | Permalink

August 4, 2006

Get Your del.icio.us Network Badge

The Yahoo Search Blog announced a new del.icio.us feature that makes it easier for people to connect and share bookmarks. The feature is named Network Badges, they basically shows others "your user name on del.icio.us with a link to your bookmarks, how many people you have in your network, how many fans you have, as well as a link others can use to easily add you to their own del.icio.us networks."

If you are interested, here is my network badge, keep in mind, I personally don't use del.icio.us all that much.

Posted by Barry Schwartz at 10:19 AM | Permalink

July 25, 2006

Yahoo Invests In Social Search Research

Reuters reports Yahoo hired Dr. Raghu Ramakrishnan as vice president and Yahoo research fellow. Dr. Ramakrishnan is a well-respected database expert who has joined Yahoo to study "links between computer and human-aided Web search." Honestly, I am excited what this can potentially mean for social search. Yahoo has so many properties that can be tightly integrated with social search; Flickr, Del.io.us, Yahoo Mail, Yahoo Video, web search, desktop search, Yahoo Groups and so on.

Posted by Barry Schwartz at 9:31 AM | Permalink

May 25, 2006

del.icio.us Adds Most Popular Links On Home Page

del.icio.us announced that they have made the del.icio.us popular page more visible by adding the content of the page, in the form of a "hot list" to the del.icio.us home page. So, now when you visit http://del.icio.us/ you will see "the del.icio.us hotlist" featured with hourly updates from the popular page. SEOs & SEMs will soon report how much of an impact this change will have in traffic and link bait retrieval, until then, it is hard to know.

Posted by Barry Schwartz at 9:11 AM | Permalink

May 12, 2006

del.icio.us Adds Inline Editing

The del.icio.us blog announced that they have added a few more features, including inline bookmark editing and prettier URLs. The changes enable users "edit and arrange your bookmarks much more quickly" and enhance the ability to "get at the conversation around the links." More details at the del.icio.us blog.

Posted by Barry Schwartz at 9:58 AM | Permalink

April 28, 2006

Your Network Added To del.icio.us - Yahoo My Web Like Features

del.icio.us announced a new feature they call "your network." You network lets you keep track of your friends most recent bookmarks. You can view your network by clicking on the "your network" link at the top of del.icio.us or by going here. How do you add friends to view their bookmarks in your network? Just click on their usernames and you will see a link at the top that reads "add [username] to your network." This is very Yahoo My Web like, is Yahoo building out two different My Webs?

Posted by Barry Schwartz at 11:08 AM | Permalink

January 17, 2006

WSJ Profiles Digg News Technology Site

Digging Out the News at the Wall Street Journal gives you some background on Digg, a two year old technology news site that's risen in popularity over the past several months.

Posted by Danny Sullivan at 8:14 AM | Permalink

January 12, 2006

TailRank: A Social News Recommendation and Filtering System Gets a New Look

Scoble points out that Tailrank a service that, "finds the best content from thousands of blogs" (think digg or Memeorandum) that provides a "custom ranking specific to the user" (think Findory) has a new look and design.

TailRank also offers a mobile version (cool!) an API and is searchable. Registered members (free) can also import blog subscriptions to build a personalized reading list. Findory's personalized reader which was released last September also allows you to import your blog subscriptions.

The TailRank home page has a feature-filled left column that allows you to focus the page by time (1 hour, 2, hours, etc.), view "hot" tags, and find links to "hot" posts.

Registration (not required unless you want more personalized results) is simple and fast. Once registered you can add tags to items you read via the service and are also given additional tools to narrow your results page. You an also add your comments to any entry.

Although the TailRank home page mentions it finds material from blogs I also quickly spotted material from mainstream news sites like MSNBC and ABC News. I also found press releases like this one from the Nikon UK site and pages from company web sites.

Finally, most TailRank entries includes an image of the page, a permalink link, a text snippet, and and a number with the total amount of inbound links to it. Clicking the "inbound link" link shows you where the links are coming from.

TailRank comes from San Francisco and is lead by Kevin Burton, a co-founder of NewsMonster and Rojo.

I'm looking forward to spending more time with the service, especially its mobile version which might be great for quick pop-ins to see what's happening in areas I'm interested in. More later.

Posted by Gary Price at 1:36 PM | Permalink

December 23, 2005

Meta Search Tagged Content With Wink

Want to meta search across content that has been specifically tagged in various ways? New search engine Wink is now live allowing that. It pulls back material categorized over at Digg, Yahoo My Web, del.iou.us, plus you can now tag things you find within Wink itself.

More is explained in the FAQ. After tag-based results come regular web search results, powered by Google. If the tagged material is lacking (or disappointing, as I've been finding them), web search serves as a backup.

By default, tagged results are sorted by "rank," though what exactly this is not explained. You can also sort by date, which is probably the way to go if you're trying to meta search for the latest tagged content on a particular topic.

Overall, I like the idea of meta tag searching because it can be a useful way to find the latest stuff being bookmarked on popular topics across various tagging communities. For Wink to be a success here, I'd like to see:

  • A full list of all the tagging communities it hits.  
  • Results sorted by date by default.  
  • Source community in results. Sometimes the URL gives you a clue where the item has come from, but not always. Take a tip from the long established meta search engines for the web and cite the originating search engines feeding items.  
  • Probably the ability to custom select the communities you want to include.  
  • Sigh. Perhaps some suggested semi-structured tags for people to use.

Posted by Danny Sullivan at 10:11 AM | Permalink

November 16, 2005

Eurekster Launches Community-Powered Search "Swickis"

Eurekster, known for its personalized social search efforts, has introduced a new service aimed at personal and small business web site owners that allows them to create a customized site search tailored to the content of the site. Creating a "swicki" for your site is both easy and free.

After signing up, you create your initial swicki by training it by responding to a set of questions that establish filters for the kind of content to include or not include in search results. The swicki then dynamically shares the most popular and active searches in that community.

Got a site? Try building your own swicki and see how it works.

Postscript: Here's one that lets you search against the SEW Blog that you can try:

Posted by Chris Sherman at 6:48 PM | Permalink

October 25, 2005

Several Social Bookmarking/Tagging/Community Search Services Reviewed

PCMag.com offers a roundup of reviews of several social bookmarking/tagging/community search services in the article: Share and Play Tag on the New Web Playground. The services discussed are:

Hard to pick a "clear winner" in these reviews. Since all services are free, try them all and see which one works best for you.

Posted by Gary Price at 1:17 AM | Permalink

See More Posts From:

This Week | This Month

  var gaJsHost = (("https:" == document.location.protocol) ? "https://ssl." : "http://www."); document.write(unescape("%3Cscript src='" + gaJsHost + "google-analytics.com/ga.js' type='text/javascript'%3E%3C/script%3E")); var pageTracker = _gat._getTracker("UA-564586-7"); pageTracker._setDomainName(".searchenginewatch.com"); pageTracker._trackPageview(); window.collarity_appid = "incmedia"; //> //>
Top Jobs
more Jobs

Account Manager
Varick Media Management New York, United States

Reporting and Data Analyst
Varick Media Management New York, United States

Director of Marketing Communications
Avery Dennison Brea, United States

Publisher
Confidential Leading Publisher New York, United States

White Papers
more Research

0