Interviews
5 Questions with PreFound.com CEO Steve Mansfield
|
Written By Reprise Media | May 24, 2006 | Share This
|
|
Today, Searchviews revives a long dormant feature - the 5 Questions interview. For the grand re-opening, we were pleased to have the chance to quiz Steve Mansfield, CEO of PreFound.com, a site that marries the functionality of traditional web search with “the human touch,” a social community that enhances the relevance of search results via tagging, rating and sharing. From his perch in ‘Silicon Holler’ (Lexington, Kentucky), Steve talks about the evolution of social search, how to handle competition from the big guys, and the discreet pleasures of the graphic novel.
1. Recently, PreFound started offering links to an external page featuring ‘traditional’ search engine results that are organized according to community behavior - a departure from the shared search results you launched the service with - and now you’re talking about integrating both types of results on the same page. Why is PreFound de-emphasizing the shared results, which seemed to be the site’s bread-and-butter when it launched?
We’re really not de-emphasizing our User Shared Search Results (USSR’s) and I would say they are still our bread-and-butter. We believe that users know what machines don’t, and that is the best way to reduce search chaff and give searchers the best results possible. But, we do want to extend the service we can offer to users by adding Traditional Search Results (TSR’s), in order to present the searcher with relevant results regardless of what they are searching for. Of course, even with TSR’s, we want to keep the search “social”, so we partnered with Eurekster to provide “community-based” ordered search results, instead of the old page ranking systems of the major search engines.
While USSR’s work great for tons of searches, they are primarily “reference” related, meaning they are the most relevant for searches looking for applicability to a subject or person. TSR’s can be less specific and non-reference in nature, making them good for searches that are looking for a single answer or something extremely obscure. We felt the best way to give our users the best possible service was to start with USSR’s, and if that doesn’t give them the results they’re looking for, then we’ll offer them community-organized TSR’s. It’s an evolution of our service.
2. Critics question whether the kind of search championed by PreFound has the capacity to ’scale’ as well as other types of search - for instance, it looks like members of geographically specific communities have the clout to skew the results in favor of relevance particular to their locations that might not be of interest to people searching elsewhere. Does PreFound have a plan to combat this tendency toward local drift, and to promote better mainstream relevance in general?
Yes, we do. The real problem isn’t that the PreFound.com concept can’t scale, it’s actually that it can and MUST scale. The larger the social network, the less of a problem skewing of results will become. We’ve always had the “chicken vs. the egg” issue, in that we are a “social search” site, meaning we need to have a fairly large social community to be good at what we do. Of course, Day One, that is fairly tough to do. So, we have always known we would have to grow fast to make ourselves truly viable and our USSR’s really relevant. In these first important months post launch (1/17/06), we’ve been successful at attracting users that are more interested in sharing what they’ve found than simply searching the PreFound.com site. Those types of users are our demographic for these first months out of the gate. As our community grows, we will start targeting more mainstream searchers who are just looking for relevant results, and not that interested in sharing what they’ve found.
This also applies to our TSR’s. Since all our TSR SERP’s are organized anonymously and directly by our PreFound.com community, it’s important to have enough users to actually affect the ordering of the results. So again, we are going to build scale to make all this a reality.
Finally, I should say that we are going to evolve even beyond that. PreFound.com will always offer USSR’s and TSR’s in combination, encompassing links to all types of media (video, audio, graphics, text links, etc.) from a single search query, instead of from different search engines like all the big sites. But, more importantly, in the next few months you will see PreFound.com offer tools that will organize these results by demographic, location, interests, participation in communities and sub-communities, etc. Users will be able to control how much or how little these elements will “weigh” the results they see from a special control panel. So, PreFound.com is moving fast to offer next-generation social search tools in order to attract the scale of users we need.
3. You’ve talked about partnering with established social networks, the assumption being that the stronger (and bigger) the community is, the better PreFound’s results will be. You recently co-hosted the Kentucky Derby Roundtable on social networks and social search with blog powerhouse Six Apart in attendance. Is a deal with something like Livejournal in the offing, and even if nothing like that is imminent, how would such a partnership play out?
We deeply appreciated Ben and Mena Trott’s participation in the Roundtable. We learned a lot about their vision, and we at PreFound.com believe that blog posts are the ultimate “social search result”. In the next few weeks I’m doing a whirlwind tour of San Francisco and Los Angeles talking to established social networks about extending their service into social search. The interest is certainly there, and there are a lot of good reasons for it. For one, social networks, like all Internet business models, have to find new and better ways of monetizing their usership. By extending into Search, these social networks can add an additional revenue stream (sponsored search links) that is tried and true. Another motivation is giving their users a search experience that is “of the people, by the people and for the people”. As I discussed earlier, a social search experience using a community of scale (for example, Live Journal has 10 million users), we can doubtless offer them search results with a relevance unheard of even a few months ago.
How it will look will depend on many factors, but imagine being on your social network site and deciding you want to do a search. Instead of leaving the property and going to Google, you instead use the web search box offered by your social network/PreFound.com. Utilizing the technology I’ve discussed earlier, you would just enter your search query in the box provided and then be presented with search results (USSR’s and TSR’s) organized by your demographics, location, interests, participation in communities and sub-communities, etc., all controlled by you, to get hyper-relevant results, all organized anonymously and directly by you and the social network you are using. It’s a whole new paradigm in Search.
4. A couple weeks back, Google debuted a couple of services that must have looked awfully familiar to PreFound, namely Co-op and Notebook. What do PreFound and its PFfinder application bring to the table that these Google products overlook, and did Google do anything that made you go ‘hmm, that’s not such a bad idea?’
We really looked at it as a validation of what we’ve been saying since January. When Google said, “Google Co-op is about sharing expertise. You can contribute your expertise and benefit when others do the same. Help other users find information more easily by creating ’subscribed links’ for your services and labeling web pages around the topics you know best,” it was like PreFound.com wrote the script.
This is validation that Social Search is the Search of the future. Now that Google has entered the space and people like Alan Eustace say out loud, “Machine algorithms aren’t good at it”, when referring to machine-based algorithms and “For the time being, the human judgment is still much better,” we here at PreFound.com have gone from being viewed as a search outsiders to looking like prophets. Virtually anything Google throws out into the marketplace is seen as brilliant, so this is going to make our work at PreFound.com exponentially easier.
It should also be noted that Google Co-op is extremely immature technology. It requires XML skills that are beyond even relatively skilled programmers. So that is not going to be competitive with PreFound.com anytime soon.
With Google Notebook, the same types of things apply. Our PFfinder tool is five years more mature than Google Notebook and does a number of things that Google Notebook doesn’t, including browsing related tools, a “fly-out” menu with three iterations in addition to commands in the right-mouse click menu, instant gathering of web page links in addition to the links on the page itself, instant emailing of gathered links, the ability to save Groups of links to a users’ personal computer and not on a web page, adding graphics to links that didn’t have them or changing the graphic of links that do, etc. Additionally, PFfinder includes patent pending technology that crawls the pages that you are saving and automatically tags them with consistent and usable tags (which users can edit). So, we feel strongly we have a mature, competitive toolkit out there.
Finally, I should also say that I believe that competition is good, as long as you’re prepared for it. If no other companies are trying to do what you are trying to do, then there is probably a reason. Competition gives you a way to contextualize your technology, so now we can say that PreFound.com has a similar vision to something Google is doing. But we have to be prepared for competition. That’s why PreFound.com didn’t bring this iteration of our technology to the public until we had two issued patents and a robust patent portfolio in our war chest. And we are extremely diligent in analyzing our competition related to patent law.
With all that said, some of things that Google and Yahoo are doing convinced us that we should start integrating TSR’s into our SERP’s sooner rather than later. We are coming around to the idea that offering both (like those guys) is a good idea.
5. PreFound is social, which by its nature is about people and their interests. Do you have any one passion in particular for which you would say, “PreFound has the most relevant result on this topic, hands down?”
That’s kind of funny, because one of my hobbies that my kids always tell me not to discuss with anyone is my interest in comics and graphic novels. They feel like I should keep my nerdy-ness to myself. Doing a search for “graphic novels” on PreFound.com will give some really great user-shared results. This one is my favorite: http://www.prefound.com/view.php?cp_id=274170
Sometimes, it just takes the human touch. And as we grow, it’s just going to get better and better.
Wow. We’d like to thank Steve for taking the time out to so thoroughly answer our 5 Questions. Interested in participating in a future 5 Questions at Searchviews? Please drop us a line.
Vinton Cerf Pushes Book Search
|
Written By Reprise Media | May 18, 2006 | Share This
|
|
Vinton Cerf might be best known as a founding father of the internet, but it was in his capacity as Google’s chief internet evangelist (and bibliophile) that he sat down with Washington Post reporter Leslie Walker to discuss Google Book Search. The point of the project is to fully digitize as many books as possible to make them searchable, which he puts into perspective from a book lover’s point of view:
“Think for a moment about the dead-tree problem…When you stand in your own personal library looking for something and you realize that A, you can’t remember which book it was in, and B, there’s no way you can go through manually looking at all the pages, then you think, ‘God, I wish all this stuff was online.’”
The publishing industry, surprisingly enough, does not exactly agree with that sentiment. Although Google does not make the full texts of copyrighted works available, it does return “snippets” of content - a few lines at most -
containing users’ search terms. Whether snippets are fair use or not is hotly disputed.
But Cerf insists that Google is out to help, not hurt, publishers. For instance, Google Book Search links to sellers like Amazon to entice users to buy the books containing the info they’re searching for. And Cerf sees a future in which a book’s inherent value would be increased by being published on the web and behaving like other web content, with reader annotations, audio and video enhancement, and links to related media.
“Because the Internet is a computing environment, a software environment, it’s possible to create a much richer kind of information than what we are typically accustomed to in books,” he says. Check out the rest of the piece here.
Interviews: Battelle Asks Lanzone, Odden Meets Whalen
|
Written By Reprise Media | April 19, 2006 | Share This
|
|
Good interviews always seem to come in twos, and John Battelle got Ask.com Senior VP & General Manager Jim Lanzone to sit for one of them. While there’s no mention of primate-fueled TV ads, there’s plenty of talk about what an “underdog” engine can offer in a search world dominated by Google’s “10 blue links.”
Battelle calls out Ask’s reported relevancy problems for last names and business names. Says Lanzone, “we’re far better than we’ve ever been, but we’re not as good as we want to be.” To that end, he says to look for gradual improvements in Ask’s algorithmic search between now and the end of the year. As for right now, he says Ask is “doing some things better than” Google, and “we feel passionately that there is more to search [than] the Google paradigm.” When Battelle points out the size disparity between Ask and the big three, Lanzone pleads for some perspective:
“…Just one point of share has an incredible impact on our business growth, and I think people forget that because they’re comparing us to Google, rather than to our own growth curve…we’re still a top player in the #1 activity online outside of email. We’re anything but small. Collectively we operate the 6th largest (just passed Amazon last month) Web property in the world. It’s just that you’re looking at us relative to the GYM behemoths, who admittedly are a level above us.”
From an SE to an SEO Elsewhere, Lee Odden has a comprehensive chat with writer, speaker and SEO consultant Jill Whalen, who’s been ranking highly since the days of Infoseek. She offers advice to SEO newbies, and really cuts loose when Odden asks what the engines could do to better communicate with SEOs. A snip:
“You’ll find I differ from most SEOs on this question. I don’t think the search engines should be communicating with the SEO community at all. Why would they? Why would we need them to? Our job is to figure them out not to have them spoon feed us what to do. What fun is it for them to tell us the rules? Blech, how boring that would be!”
Related: Our own sit-down with Lanzone, from way back when a guy named Jeeves worked with him.
Google CEO Steps to the Mike; Battelle Quizzes MSN Live Labs Man
|
Written By Reprise Media | April 7, 2006 | Share This
|
|
Top men in search are letting loose with words this week. Google CEO Eric Schmidt made a speech last night to the Economic Club of Chicago, recapped in a post at VC Confidential (via Threadwatch). And John Battelle corralls a refreshingly candid Gary Flake (formerly of Overture and Yahoo!), the founder and director of MSN’s Live Labs.
Schmidt used his speil to stress the importance of targeted advertising in a world populated by the ever-distracted, and that untargeted styles of marketing used in media like TV and telephones would be increasingly irrelevant as consumers use the likes of TiVo and ‘do not call’ lists to filter out unwanted ads. Instead, he touted the consumers comprising online social communites as an essential marketing goal. Schmidt also noted that the collective knowledge of such groups is also largely untapped; he spoke about attempts to mine the internet’s wisdom of crowds, including their predictive abilities - for instance, certain hedge funds have been delving into chat rooms to suss out stock assessments.
Over on Battelle’s blog, it didn’t take much coaxing to get Gary Flake to open up. He talked about search architecture (”64 bit systems pave the way for entirely new forms of relevance that look at how pages relate to one another”), owned up to MSN’s past mistakes (”What didn’t I like about our old UI? Where do I begin? There is so much to choose from…to be blunt, it sucked”), and the pros and cons of Microsoft’s culture. Flake’s not shy about cheerleading for his colleagues (”these are simply wonderful people in every way”), nor about taking a hard, self-critical look about how MS operates:
“I don’t think MS is particularly clear in how it communicates things to the outside world. I also think that MS often makes mistakes by trying - wait for it - too hard. Specifically, instead of doing one thing and nailing it, we’ll sometimes do six slightly duplicative things in parallel - each being slightly below critical mass - which can create confusion inside and outside of MS until we’ve had time to sort things out better. So, yes, I get frustrated at times. So it goes. For me, frustration is a biological signal that means I should try harder, which is a good and helpful thing.”
Related: MSN Mad About Labs
Lessig Talks DRM with Lenssen
|
Written By Reprise Media | April 3, 2006 | Share This
|
|
April Fools day is passed, with its attendent web tomfoolery in the can (Wikipedians compiled an exhaustive list of pranks). Even if you’re only a little sick of Saturday’s leftover hype, Google Blogoscoped has the perfect cure for a high jinks hangover: Philipp Lenssen distracts Stanford Law Professor and copyright reform advocate Lawrence Lessig from an extremely busy schedule (”too much travel, too much email, not enough time for reading or writing”) to summarize his views on the problems posed by Digital Rights Management (DRM), spam blogs and DVD region codes.
While Lessig advocates a simple solution to the fair use restrictions created by DRM in software, DVDs and other digital content - “allow [users] to turn the DRM off, and use the work under an assertion that the use was fair” - it’s a wholly unlikely outcome in the short term. Lessig explains:
“The problem is fear. The copyright industry fears the internet, and they are exercising their (considerable) power to achieve control. That effort is creating huge problems for a wholly creative and productive use of digital technology.”
Lessig dislikes DRM in part because he’s uncomfortable with the idea of automated policing of copyrights, since “copyright was designed with a human enforcer in mind.” He concedes that although human enforcement is a “difficult” proposition, “repressing speech” - a consequence of DRM impeding fair use - “shouldn’t be easy.” What about the spreading problem of spam blogs, which copy content wholesale without attribution or permission in order to generate ad revenue? Lessig suggests getting ordinary people involved via “legislation and a bounty system to make it enforceable,” an idea he’s explored in further detail in a previous column.
The importance of citizen involvement in pushing for fairer copyright laws is a recurring theme in the interview. Lessig’s morally opposed to law-breaking (that includes illicit peer-to-peer file sharing), even if it’s a law he disagrees with - except if it constitutes civil disobedience. Calling copyright law “insanely complicated,” he urges people to “demand the regulators justify their control.”
Robert Scoble Interview
|
Written By Reprise Media | March 31, 2006 | Share This
|
|
Do you read SEOBuzzBox? If not, then shame on you.
It’s a great blog penned by Aaron Pratt, a rain barrel craftsman who likes to talk about marketing on the net. It gets no more niche than that. Well, it probably does, but humor us, OK? Read more.
Today Pratt’s posted a great interview with Robert Scoble of Scobleizer fame.
Here’s an excerpt in which Scoble talks about doing it all for the kids:
“PRATT: Some do not understand these PR spammers who visit your blog and accuse you of being an elitist. I also have never understood Microsoft haters in general, is this just part of working for a big company? Have you ever considered banning these people from commenting in your blog or would this restrict the ‘conversation’ as you and Shel Israel call it?
SCOBLE: Yeah, I think it’s part of belonging to a big company and it’s also part of being popular. American culture loves to build people (or things) up and then tear them down. Heck, I root against the Yankees for the same reason. Yeah, I have considered blocking comments. I think at this point, though, that the ongoing conversation is more valuable than the emotional pain from having people say mean things to me.
Awwwwwwwww…..
BTW, this is probably older than dirt, but we also happened upon this Robert Scoble trading card. Scary.
SearchViews “5 Questions” Best Of
|
Written By Reprise Media | March 27, 2006 | Share This
|
|
Ah, the sweet smell of five. From time to time we at SearchViews sit down and talk to the people making news in the world of search and tech.
Here are some of our favorites:
Seth Goldstein This co-founder and chairman of Majestic Research likes using complicated Star Wars analgoies to explain the changing role of the agency as it relates to search. Find out what the heck he’s talking about here.
Peter Pezaris This guy took that whole “Go forth and…” edict literally with a social networking service that’s topped a million members. Read our interview here.
David Jakubowki MSN Manager of Search Strategies and Go-to-Market sounds off on his feelings about the SEM/portal divide and the impact of demographic targeting on overall search spend.
Jim Lanzone He’s the senior vice president of search properties at Ask (Jeeves) who knows his way around vertical search and every celeb blogroll under the sun.
David Zito He’s senior manager of development for Yahoo Publisher Network Online and a start-up guru with a track record of success. Find out what he’s doing for the big Y over here.
“And Now, the Thrilling Conclusion”…Andrei Broder Interview Pt. III
|
Written By Reprise Media | March 16, 2006 | Share This
|
|
Yesterday, the Yahoo! Search Blog posted the last of their three-part interview series with Yahoo! Research Fellow/VP of Emerging Search Technology/All-Around Search Maven Andrei Broder (we covered part 2 here). The trilogy winds down as Broder answers questions from YSB readers, including a couple of interesting ones about spam and blog search.
Broder was first asked how he felt about Alta Vista’s collapse. He returned a curiously Vulcan-esque response, describing what happened rather than his emotional reaction to what happened. He said that the erstwhile search leader (and his old stomping grounds) suffered from “almost perfect bad timing,” squandering its technology advantage while operating a bad business model. ‘Yes, Spock, but how did it make you feel?’ Maybe it’s still too soon.
He was on more familiar footing while discussing blog search, and in particular why it’s not that great:
“If you look at web search in general, the biggest help comes from metadata, anchor text, links, web graph analysis, etc. For blogs we have very little useful metadata. And even if you do have metadata for blogs, it is often wrong, or you can’t trust it, so you use it very little.
Broder also argued that blogs are often contextually confusing, for search engines and other entities. “Even a human doesn’t understand what’s going in a blog if dropped into the middle of it.” (Do we smell a Tron remake for the blog age? …Sorry, our nerd is showing, and we digress).
Finally, the answer to quashing spam? Make it too expensive. Even if you delete every de…um, ’shmebt shmonsolidation’ email you receive, enough people are clicking on them to make the business worthwhile. “…people think spammers are kids up to no good, but it is not. Spam is about economics.” This seems to speak to Yahoo!’s recent plan
to charge email marketers a fee for a guaranteed email delivery. But Broder thinks that with the increasing personalization of search, spam could become more difficult as a rule since it’s “hard to make robots that behave as humans.” A fine observation, but spammers are people, too; we’re afraid they’ll have not enough difficulty adjusting to the personal age of search.
The Future of Search? Toss the Box: Broder Interview Pt. II
|
Written By Reprise Media | March 10, 2006 | Share This
|
|
The Yahoo! Search Blog posted a continuation of their Andrei Broder interview series yesterday (installment 2 of 3 - we gave part 1 a shoutout here), in which the VP of emerging technologies expounds on Dilbert cartoons, the improvements since the Jurassic period of search - a time when Alta Vista’s Tyrannosaur ruled the landscape - and the industry’s ‘next phase,’ which he describes as “search without a box.”
Citing GPS-equipped cars as an example - automotive references are a recurring motif - Broder imagines that search will soon provide more and more information “in a context without actively searching,” in effect “going from 2.7 words per query to 0!” Today’s navigation systems, for instance, show you the locations of gas stations along your route, and travel sites return weather reports and hotel suggestions for destination locales. A couple of years from now, he hopes, the information would be provided in “less constrained” contexts, and on an as-needed bases - telling drivers the locations of gas stations only when the car is low on gas, for instance. He says that search will “move from information retrieval to information supply.”
This speaks to the trend toward search personalization; obviously search engines aren’t going to be able to read minds or tell the future, but in order to predict users’ (or a “class of equivalent users’”) wants and needs, engines will require more pertinent information, and more of that information - whether it comes from your Honda’s gas tank or your Yahoo! Shopping search - will have to be personal, if not uncomfortably intimate. Broder does note that in terms of the information supply, “there is a fine line between annoying and useful.” We would hope engines apply this warning to information retrieval as well.
What’s interesting is that Broder seems to think that in the transition to the next phase, search should take its cues from advertisers, because they rely as much as anyone on the abilty to place information in the most relevant contexts. “…Advertising is a form of contextual information supply…Information supply as a science will continue to grow because of advertising.”
Tech Talking Heads
|
Written By Reprise Media | February 22, 2006 | Share This
|
|
If only it were Thursday instead of Wednesday then we’d win some sort of award for alliteration with this Wednesday ticket of three tech talking heads. Oh well, there’s always next week…
Interview #1 The Wikipedia Signpost is a “community-written and community-edited newspaper” covering topics and events of interest to the hardcore Wikipedia set. This week it features a great interview with Wikipedia founder Jimmy Wales. All the questions were submitted by Wikipedia users and fans. Read it here.
Interview #2 The service formerly known as TiVo may be getting an extreme makeover. The box + subscription model could soon give way to a subscription only model, similar to what cable companies offer with their services. TiVo CEO Tom Rogers explains in this interview with Bloomberg News.
Interview #3 Last week he was the one being interviewed, this week he’s doing the interviewing. We’re talking about Lee Odden, turning the tables with this interview of Traffick.com blogger and SEM guru Andrew Goodman.