Written on April 18, 2008 – 8:30 am
Ernst-Jan Pfauth, editor in chief
The Next Web Blog covers start-up news from all over the world (not just the Valley), exciting new technologies and inspiring entrepreneurs. If you're new here, you may want to read our '
About' page and subscribe to our
RSS feed.
If you came here from Digg it would be great if you could actually Digg us too! Do you have a start-up that we should write about?
Contact us! Thanks for visiting and hope you come back again!
If you follow blogs like The Next Web and Altsearchengines, you might notice that social search engines are booming. Stumpedia, Topicle, AndUnite, you name it. They’re popping up everywhere and at least twice a week, one of the founders sends me a line to tell me why his engine is gonna be the top of the bill.
Geoffrey McCaleb from Nsyght is one of them. He and his team have found a new take on social searching: allowing users to use their bookmarks and their social graph to create a customized search experience. “the real eureka moment came when I thought, how can an algorithm tell me whats relevant when my assumptions of relevancy are going to be different from everyone else.”
McCaleb believes that no algorithm can replace the objectivity of a human being. “We all have different concepts and notions of what we find relevant. So, we wanted to create a search engine that didn’t treat every set of keywords the same. What you find relevant for a search may be different from mine.”
The beta version of Nsyght provides integrations into a number of social services such as del.icio.us, ma.gnolia, simpy, digg, last.fm, twitter, and pownce. “We do a couple of things with the social graph, we preserve friendships across different social networks, and we allow for users to syndicate their bookmarks between these services. The key takeaway point here, the more information we know about a person, the more we can customize the search to make it more relevant to them.”
The holy grail
But the most important question remains, Mr. McCaleb. Why would the public choose your search engine, out of all the other alternatives?
“Well there may be a lot of competition in this space, but I feel pretty strongly no one has approached the problem in the same way we have. The holy grail in search is not having the biggest index, but the one that gets you a relevant result in the shortest possible time. While we are small - a given since we are self-funded - we obviously have a ways to go. But as we gain more users, we will gain more of their bookmarks, and then we will have even more highly relevant sites in which to crawl and index.
Another way to look at it is this. PageRank is a brilliant concept, and still does an incredible job determining relevancy. But fundamentally, even with all the data points probably looks at it still is more concerned about the source and not the content. What we can do is let the user define the sites they feel are relevant, leverage their social network, and over time see their results become highly personalized. So in a way, augmenting the algorithm for their own use and gain.”
What do you use?
The only way to end this article, in my opinion, is to ask you guys whether you really use social search engines? Since you’re probably the most web-savvy crowd out there, you would also be the first ones to adopt a new trend. So tell me, still on Google? Or already switching to the brand new social engines?
Written on April 7, 2008 – 11:15 am
Ernst-Jan Pfauth, editor in chief
Every week we publish an interview with a start-up. We ask five questions, hoping the answers will give you inspiration and new views. Well, actually six questions, since we also ask the start-up to who he or she is passing the mic to.
This time we’re interviewing Luis Pereira, Founder & CEO Stumpedia, a human-powered social search engine that enables registered users to submit sites and matching keywords and phrases. The relevancy of search results are then ranked and rated by the community. The future is search is social, said StumbleUpon founder Garrett Camp last week during The Next Web conference, so I figured I should ask Luis to participate in this start-up series. Want to know exactly how it works? Check out this extensive article on SearchRank. Want to know why Luis started this search service and the influence he expect it to has on the future web? Then continue reading:
How did you come up with the idea of Stumpedia?
“The initial idea for Stumpedia.com came as a result of another web site we launched called AskPoodle.com. Both web sites are described as being human-powered, where as Stumpedia.com is a global search engine for key-word driven searches, AskPoodle.com is a local business directory for the US market. We are in the middle of a fundamental shift in the search space. Page Rank will certainly be around for a very long time, but how people search and surf the web is changing. Social bookmarking sites are a great example of new search habits and trends that are emerging. We recognize a need in the market for web pages and web sites to be ranked based on social collaboration methods. Social bookmarking data such as tags are one important element in determining search relevancy.” (more…)
Written on March 8, 2008 – 12:03 am
Ernst-Jan Pfauth, editor in chief
These days everybody is talking about the ‘open’ trend on the web. So do I, since it’s huge. Yet I also notice another trend: finding similar persons. It looks like people finally discovered that the web is an ideal medium to meet like-minded people. Last week I discussed matching service Lovest.at, today it’s all about a German service that turn search engines into meetingplaces: andUnite.
The service launches in an English and German version tonight, and by then you’ll be able to find people with similar interests, based on your search terms. You can sign up (anonymously), install a plug-in that creates a social overlay on Google & Co and the service automatically starts collecting your search terms. The advantage for the founders is obvious: valuable information. Yet what’s in it for us?
I’ve asked co-founder Bernd Storm van‘s Gravesande: “Christian Schmidkonz and me were both frustrated with the boring and lonely process of web search. We thought that it would be always more useful to be able to ask someone who knows something about a problem or question instead of browsing though pages and pages of more or less interesting search result links. We thought that the value of a search term must be much higher than just being used for retrieving links from a database.”
Bernd and Christian were also curious what happened when they’d collect search terms in a profile: “Would the collection really represent our interests or even characters? And would we be able to get to know interesting people if we match our profile with the profiles of others? So, basically we started with an experiment but already with the first prototype it turned out to be a intriguing and addictive tool we created.”
“Most people don’t want to change their favorite search engine.”
They hired Christoph Fuchs, a ‘highly talented’ student, to help them developing the service. Now, it’s all about promoting their special concept. “We have to convince the users every day, that andUNITE offers a unique service which is not only useful but also fascinating and as already mentioned addictive.”
AndUnite immediately reminded me of Jimmy Wales’ plans to build a Google-killing social search engine. So far, people don’t seem to be really excited about Wikia Search. Why would andUnite do any better? Bernd: “Well, Jimmy Wales tries to set up a completely new search engine where users have to participate actively in the so-called social part of Wikia in order to make it lively. With andUNITE we sit on top of Google, Yahoo! or Live. We think that this approach will be more successful because most people don’t want to change their favorite search engine.” (more…)
Written on February 22, 2008 – 1:18 pm
Ernst-Jan Pfauth, editor in chief
Mobile internet is hot - we all know that - and one of the rising stars in the mobile world is Taptu, a social mobile search engine. They won the Global Community Award at the MoMo Peer Awards on February 13 at the Mobile World Congress and Robert Scoble proved on PodTech that it comes up with better results than Google does. Officiously, it seems like Taptu has the potential to become a big player in mobile search. So it’s about time that we hear the story behind this UK-based start-up. After a nice and interesting email conversation with the Taptu team, CEO Steve Ives told us where he got the guts from to challenge Google and what the social aspect of Taptu is.
First of all, the Google Question. Every time a new search engine sprouts, this is the first question critics try to answer. Ives: “A Nokia guy I met at a party confided to me that he thought Google still hadn’t yet cracked mobile search properly, and some fresh thinking was needed in this market. We started to look at it, and the more we looked, the more ideas we came up with. It became a bit of an obsession after a while. We pestered the VCs constantly after that, eventually they gave in and invested in us.”
When moving a service to mobile, something gets lost in the translation.
Google’s position seems untouchable when it comes to desktop search, but challenging the giant on the mobile phone might work. Ives explains why: “Services like Google were born on the desktop and then moved later to mobile. When moving the service to mobile, something gets lost in the translation. A desktop user will use search 5 times a day or more, but a mobile user that discovers Google Mobile or Yahoo OneSearch typically only searches once every 5 to 7 days. We believe that to get people to use mobile search 5 times a day or more - in other words, to make mobile search a mass market service rather than a niche service - then you have to give it a social context. Mobiles are supersocial devices, so if your service isn’t relevant to you in a social way it won’t get used that often.”
Next to a Google-challenger, Taptu is also a social search engine. Jimmy Wales plugged this term in January when he launched Wikia Search. During an interview I had with him, he said: “One of the weaknesses of current search engines is that their algorithms take a long time picking up new good sites. (..) It takes only one community member that finds a good new site and lets the community know.” What does Ives thinks of Wikia Search? “We’ve been watching Search Wikia quite closely. They have some similar ideas to us about how to improve the quality of search results through social interactions. But we are totally focused on mobile, which is a completely different medium to the desktop, so the dynamics of sharing are rather different also.” (more…)
Written on January 7, 2008 – 9:59 pm
Ernst-Jan Pfauth, editor in chief
After waiting for a year, it finally happened today: Jimmy Wales launched an alpha version of Wikia Search. After such a long wait, people have high expectations. Unfortunately, Wales didn’t manage to live up to them. TechCrunch’s Micheal Arrington called it ‘one of the biggest disappointments I’ve had the displeasure of reviewing’ and ‘an inexcusable waste of time’. Stan Schroeder from Mashable wasn’t very pleased either: “Wikia Search looks like something that was cooked by two guys in one month in a basement”.
“We’re not at all pretending this is Google-quality yet, it’s far from it.”
We interviewed Wales last weekend, and he sounded a bit like he knew what was coming. Wales: “The social network we’ll launch on Monday is just a project to build a search engine. We’re not at all pretending this is Google-quality yet, it’s far from it. This is just the beginning. We expect it will take at least two years before we have Google- and Yahoo quality.”
When that time comes, Wales expects people will switch to his engine: “It’s very easy for people to switch from one search engine to another. If we do a good job I’m not too worried that they will switch to ours.”
The positive attitude of Wales has been rewarded before, yet it’s still a huge challenge to compete with Google. Doesn’t it need some hash promotion techniques? When suggesting that Wikia Search could take advantage of the Wikipedia pages always showing up in the top results, Wales answers determined: “No no, Wikipedia has absolutely nothing to do with Wikia Search”.
“I’ll use the same marketing plan as I had for Wikipedia: do a good job and people will find you.”
With that in mind, it sounds even harder to reach the audience. Wales however, doesn’t seem to worry about it: “I’ll use the same marketing plan as I had for Wikipedia: do a good job and people will find you.”
I must admit though, that the plans of Wales and his team sound revolutionary. For instance, by keeping the code of Wikia Search open source, they give other search services and organizations the opportunity to create the perfect search engine together. This sounds logical when you take in account that Wales thinks search won’t be competitive element anymore. Wales: “Good quality search is becoming a commodity item. The search quality of Google, Yahoo and Ask are actually very similar. So the idea that Google is some kind of technological powerhouse, is actually not longer true.”
Bringing the social aspect into search successfully is something we haven’t seen yet. Though the social network they have on-line now isn’t very spectacular, the promises for the future sound good. “One of the weaknesses of current search engines is that their algorithms take a long time picking up new good sites. If you look at the way Google ranks sites, it all depends on the number of important sites that link to you. In our project, it takes only one community member that finds a good new site and lets the community know. That will affect the ranking immediately.”
But what happens if a large company tells its 500 employees to give their corporate site a ‘thumbs up’ on the Wikia search engine? Isn’t that a big threat for the validity of the search results? Wales: “That remark is very similar to questions that people would have about users doing bad things on Wikipedia. It’s very difficult to fool a community. Ranking a search result is a public act, so people can see what you’re doing and will rank the contribution very low.”
“We would be thrilled if we eventually have a market share of 5 percent”
Sounds like Wales is actually planning on competing with Google. “As with every open source project, we will have a high number of languages covered. We would be thrilled if we eventually have a market share of 5 percent.” Did he receive any reactions from Google HQ? Wales: “I see the Google guys socially from time to time and told them about the search plans. But they didn’t give a real reaction yet.”
Email
Twitter
Facebook
Netlog
Linkedin
The Next Web Blog is closely associated with The Next Web Conference which is held annually in Amsterdam, The Netherlands. At this event speakers from all over the world come together to talk about, and show of, the future of the Web. (
