Silicon Valley WebGuild: Vertical Search Trends

We’ve got a great event tonight, and I would like to welcome you all once again to Google Flex, and sorry for the last minute change of venue; but looks like you’re probably here — made your way here. I would like to thank our distinguished panelists, and before I begin I’d like to make some short announcements on the WebGuild. Okay, — what is a WebGuild; who are we; and what do we do? The WebGuild is the largest professional web organization in Silicon Valley. We focus on three core areas, usability, development networks, SEO, and SEM markets. We have over 5,000 people on our mailing list, and through affiliates we reach about 25,000 people. About 42% of our members are CXOs, VPs; 41% are technical; 10% are NORS or self-employed — self-employed. The concentrations of the companies they make up are, by, from top to bottom: Yahoo, Apple, Google, Microsoft, Cisco, Verisign and HP. Sounds like a recruiters dream right there. — [Laughter] So, having said that, if we quickly look at our website, it explains our three core areas of expertise, which are usability, development network and design. And under usability Jerad Spool, who is a guru in usability, heads that group; he is an advisor to us. The development network is headed by Bill Scott of Yahoo, who runs their — one of the people who run their development network. And the Internet marketing group is going to be run by from somebody from eBay. I just spoke today with Rajiv Dutta who is the President of PayPal, PayPal eBay, and they’ll be helping us with some resources in that area. So, why is that important? Because, based on these three core areas — that is how we come up with our topics. So we find themes that are operating in these three core areas; find companies that are executing on those themes and find speakers around those themes. In this case, it was Vertical Search, so we have the; we have Simply Hired, and we have, companies that are really disrupting the market place. So, next month we have Social Search, and we have Dig, Delicious, and Stumble Upon. And the following month, we have a conference; I’ll get into that a little later. So, having said that, if you have any topics or any speakers that you are interested, please go to the website, fill out the form, and we’ll do our best to find a speaker in that area; we’ll try and have it covered. I would like to get into a little bit about the conference. The conference is on October, and the theme of the conference is the “New Web”. The focus, again, will be user and the user experience, and having said that, we are – the keynote speaker is going to be Ram Shriram. Ram was one of the Executives at Netscape and Amazon, and is one of the founding principles of Google. So he has a lot of insight into building consumer Internet companies, and he is going to be sharing some of his recipe, in fact, his secret sauce and ingredients of what he is looking for in doing that. The other note — keynote speaker is, Marissa Mayer, who is Google’s guru of user experience, so she’s going to be talking about what they do to measure the user experience and how they roll out new products across that area. And then it is going to break out into two tracks. The first track is going to be talking about technologies, the methodologies, and the platforms into which the user experience is going, such as – And Jared Spool is going to be talking about usability — I’m sorry, Yahoo and Microsoft and a few others are going to be talking about AJAX and other tools that are building the rich Internet applications. And finally Kelly Goto is going to be talking about the mobile web. And the other, track B is going to be talking about how you monetize this user group once you build it and then you add platforms. So before we wrap-up, I’ll turn you over to Reshma who is going to make the introductions. Next week or early — later this week, you’ll get a survey from us and we would like to know how we are doing. We would like to know how your membership is working for you? So, before we get into that, I would like to give you some idea of where we fit in the landscape. So some of the groups that we compete with in the events, monthly events area, as SDForum; SDForum has an annual fee of $125.00, each event is $15.00 – $30.00, and their focus is basically everything. And there is SVASE, they are about $100.00, and each event is between $45.00 to $60.00, and their focus, I think, is to be like SDForum. And — [Laughter] — then there is TiE, and TiE’s average monthly fee — sorry annual fee — is about $100.00, and each event is about $25.00 to $55.00. And, then there is the WebGuild, and we charge $65.00; that’s basically a tank of gas these days. And — [Laughter] — each event costs about $5.00, and that is basically the cost, less than the cost of a Big Mac and a small fries and a medium coke, I guess, you know. — [Laughter] So I think we deliver excellent value in this space. So why have we been able to go from the bottom to the top and attract keynote speakers is because we focus on one thing — that is the web and the three core areas in the web. So we’re moving now is into the conference arena, so we have a conference coming up. Our conference is $135.00; it’s going to be jam-packed; it’s going to be super. It is the best minds out on the web, and you are going to love it. And we compete with Search Engine Strategies (SES), which it charges approximately $1,500.00, AdTech around $1,500.00, AJAX World around $1,500.00, and O’Reilly $3,200.00. By the way, John Battle, who runs O’Reilly Network is helping me with some of his speakers, so that is why we have been able to get a huge roster, so. But having said that, we are not really competing with any of them directly. We’re taking the top areas, where they’re focusing on the web, and that’s why you should come; and we are non-profit, so we are not in the business to make money. We are in the business to deliver value to you, and if you get value you come back here. Great, so that was my – I’m sorry. When is it? Oh, October 19th, it will be on our website very shortly. Also, by the end of this week, we will have our jobs board launched, so again, it’s going to be a very focused jobs board, all Internet focused. We’re going to be giving Craig’s List a real run for their money. [Laughter] And, the format of tonight’s presentation, after Reshma’s introductions we’ll be having a discussion for about an hour and a half, followed by Q&A. Since we don’t have a mic, I think some of you will have to come forward and ask a question. This mic apparently does work or doesn’t work? It does, okay. There will be a conclusion, a Podcast of this event will be made available on the website within 24 hours, and the speakers will make their presentations available at their discretion. I would also like to ask you to turn off your cell phones and pagers, as a courtesy to other members. So without further ado, let me turn you over to Reshma, who is going to introduce tonight’s speakers. Thank you very much. — [Applause] You probably won’t hear me at all. So I will be really brief. So tonight we have three — core speakers and one moderator, and I will give you just a brief background about each of our speakers and moderator. Michael Yang is the Founder, President and CEO at Michael is passionate about developing services that help people; he is the Co-Founder of mySimon and later sold that company to CNET. Become is Michael’s latest venture; the company helps people make better shopping decisions, and since it’s launch in 2005, has become one of the fastest growing shopping sites. Suranga Chandratillake, Founder and CTO of Blinkx, and forgive me if I butcher anyone’s names, and Suranga is the Founder and CTO of Blinkx. He has more than 7 years of frontline experience as a technology innovator. He is – as Blinkx’s CTO, he has developed the world’s most comprehensive video search. And our third speaker, Gautam Godhwani, is President of Simply Hired. Gautam has founded three ventures in the last eight years; he is well experienced in starting startups and finding great people to grow them. He is currently CEO and Co-Founder of Simply Hired, a search engine for jobs. And our moderator, Sapna Satagopan, is a research analyst at Jupiter Research, where she is focused on online search covering search engine marketing, SEO, and the overall search experience. Prior to joining Jupiter Research, Sapna worked at a consultant at PWC. So it is my great pleasure to introduce to you, all of our distinguished speakers this evening. Thank you. — [Applause] I guess I am the third kind of speaker; I start off pretty loud then it goes down a bit. So yes like, Reshma said, I cover search and search engine marketing for Jupiter Research and it has obviously started to spread out in a lot of other zones, and vertical search is definitely something that we are focusing on a lot more in the recent times. And obviously one of the hot beds for that is where we are at right now, Google. Essentially what we are covering is the main general search engines, but vertical search is something that I think I’ve spoken with each one of the panelists at some point in time to discuss that, so I think it should be an interesting session. I just wanted to set a base to the discussion in terms of understanding where – why we are so excited at our research firm about search engine marketing or online search, and I promise this is the only data slide that I would put in here. I just have a couple of slides, but this is precisely the reason we think search is getting more and more interesting. This is the advertising forecast, and bear with us, so we are looking at about 40% of an advertiser’s budget that is going to be focused on search by the year 2011. And that is exciting because the growth has been phenomenal; the marketer’s interest in it has been phenomenal; and it’s easy to understand why it’s getting so big. A couple of reasons for sure; lots of people go into search engines, whether it’s going to be Google, Yahoo, any of these search engines, and the marketers are starting to notice that. There are lots more people coming in to put their ads on it; to put their, whether it is going to be Adverts or Yahoos ads to be able to reach these searchers, and that is essentially the main reason it is getting pushed up a lot more. And one more interesting aspect to that is — or the ‘push factor’ to why search is growing so huge for the advertisers. It is also the reason that they’re opening up to a new set of search engines and not just focusing on Google and Yahoo. And Michael probably knows this a lot; if I am a retail marketer, I’m selling some kind of shoes and I know I go to Google and put my ads there to be able to access the people searching for shoes. I also know that they are these bunch of other vertical shopping search engines,, a bunch of other shopping search engines that I should also be focusing on and that is definitely another contributing factor to that. And one, and this is just sort of a snippet of why I think it’s going to be interesting in retail for sure. This is a question we ask consumers — how useful they find any of these types of content in trying to purchase something? And I sort of highlighted the shopping comparison sites,,, which is starting to really rise in how useful it is and how our searchers are basically using it. Search Engines is right on top, but it’s – and it’s obvious that the top right is where the most of these content providers would want to be at; but it is an interesting point to note because shopping search engines have actually gone up a lot more on the X axis and that is definitely an encouraging trend. One more example – health search is an emerging information destination. This was a question posed again to health consumers on how they went about looking for information about health and wellness and about 60% of them used general search engines to look for information about it; 42% used search engines or sites that focused on health. So the blues — government aided engines, Seattle Health, AOL Health, — are where people are going to, to be able to look for this information. So essentially, the bottom line is — there’s — they are sort of the initiators of the market; a few of these industries are sort of poised to be able to take advantage of search markets – search engines or vertical sites by virtue of the fact that they have been around online a lot more. Retail is one example; travel is one example — is another travel search engine; financial services is something which we are starting to see a lot more interest, especially from the marketers; media and entertainment is another area which is well poised to take advantage of vertical search, because they have been online for quite a while now. The emerging categories, and this is where the interesting part really starts up — classifieds, health; there are health search engines. I believe is one of them. Real estate, there are a couple of real estate, I think search engines from the Bay area itself; jobs – which is what Simply Hired is focusing on; automotive is something that we have heard a lot of interest from the marketers and it’s definitely an area that we think either it blends into classifieds or it sort of takes off on its own. But essentially, the general search engines are starting to create these graduation grounds for each of these vertical sites. When they get bigger, when they get more and more of the volume from the searchers, it becomes an interesting point where the motivation to start a vertical site starts. So, with that I know we had a quick introduction to each of the panelists; but I would like to hear from Michael, Gautam, and Suranga Chandratillake, which I can actually pronounce because I do watch Crooked and I know there are lots of names like that in there. I would like to sort of get a sense of what you do with, Simply Hired, and Blinkx and where your sort of companies started off. So good evening, my name is Michael Yang, and I’m very happy to be here. I wanted to start by asking a quick question and see a show of hands here. How many of you have had an experience where you were looking to buy something, so you wanted to kind of get some more information, reviews or products and you could not really find that information on the Internet, whether you were using search engines or any other sites; can you raise your hand? Okay, it is about 50% — 60%, and this is one of the reasons why we decided to develop a vertical search engine that specialized in product and shopping information. So, just a brief background, we — my partner and I started mySimon back in 1998; it was one of the first online comparison-shopping sites. And we got back together about three years ago to do something that is very, very different and new in going after some problems that the general search engines weren’t really delivering. So, let me start by giving you a quick overview of our website, and then I have a few slides that gives you a little bit more background information about the company. So, if you go to, one thing you will notice that we have these suggestions at the bottom, but as you are typing something in, let’s say you are looking to get some more information about BMW. You see a lot of suggestions here like BMW, 3-Series 567. These are information that we gather from crawling billions of pages of web documents, so we are able to come up with intelligent suggestions to help you narrow down your searches or find products that you didn’t think about, but you are actually interested in. So, let’s look for something like plasma TV and type in a return. One thing I want to point out is that we have two buttons, research and shop. Basically, Research allows you to find all the great information like reviews and buying guides to get more information about a product like whether it’s an MP3 player, refrigerators, or any kind of product, and then Shop is when you are ready to shop for products and you want to compare prices, and that’s what you push. So, a default here is we give you a research results on the left-hand side, and shop results on the right-hand side. So, here we find you information from consumer research, CNET on TV buying guide; so these are very expert buying guide information on products. And if you are trying to understand the difference between plasma TV and LCD and High Definition, this is a great source of sites for that type of information. And once you’re ready to — another thing that we have that is fairly unique is this Search Zoom feature, where you can narrow down the results by buying guides or product reviews, so you don’t have to flip through so many pages to get to what you are looking for. So this is a real nice feature; we use extensive machine learning technology to make this possible. You can narrow it down to product reviews or buying guides here, and once you are ready to shop, you can click on the Shop button. And we show you many different product brands and models and their prices, and you can narrow it down by particular brand or price or by particular online store. So you could click on one of the brands that you might be interested in and then, like for this Pioneer 42″ PDP Model; you could compare prices, and we find merchants that sell that product. So we’re here in Mountain View, so we just type in the zip code to calculate the total cost including shipping and tax, and we give you the total price at the bottom. So one place sells it for $1600.00, another place sells it for $2900.00. So it’s a pretty wide gap between what some of these merchants are selling the identical product for. Another feature that we have is, — let’s say you are looking for iPod Nano. Let’s say you are looking for this 4MB, 4GB iPod Nano, Black. What you see here on the right-hand side is a nearest store; so once we understand the location of a user, then we can find the stores nearby that person to help you find – so in this case, there is a Circuit City and an Apple Store in Palo Alto and Mountain View areas. So you can actually click on this to see where the store is located. We use Yahoo Maps here with the dynamic zooming capability so that you can see the stores’ address, phone number, store hours, and even driving directions. So it makes it very easy for people to compare buying online versus offline, if you want to go down and actually take a look at the product and see how it feels and make a decision. And a lot of times, offline stores will match the online prices, so you can actually print out this result and bring it down to the store and they will usually match it for you. The last feature that I want to point out is this ‘Add to my Favorite’ feature. So if you are thinking about – if you are considering a few different choices, then you can actually put those up on the workspace at the top, so that you can think about it; look at them side by side; or you can add other kinds of products. Let’s say you are looking for an iPod Docking Station to go along with that, and you want to take a look at how they would look together, or it could be a tee shirt and a pants, or even a – any kind of product that you could think about. You could put them up on the workstation and this information is stored for you based on the cookies, so you could come back later and make your decision; and we are expanding on that, so. That is a brief tour of our website; it combines the best of search engine that finds reviews and buying guides, and we help compare prices on thousands of products. So, just a few slides on the company. The opportunity that we are going after — Is — as Sapna pointed out earlier, Search Engine is incredibly powerful; general search and vertical search combined is about a $15 billion dollar business this year, growing to $30 billion by 2010; that is according to Piper Jeffreys, our Research Report. And a lot of the monetization is through the shopping and commercial related queries, and today currently — approximately $350 million queries are done that are shopping oriented. So there are lots of people looking for different kinds of products that they want to buy or they want to research. So, the general search engines are very useful if you are looking up any kind of information about any topic; but a lot of times they are off topic when you are looking for a particular product reviews or buying guides because, let’s say you are looking for a television. It brings back information about the history of television or television station websites; it’s just hodgepodge. But a vertical search engine gives you just product information like buying guides and reviews and discussion forums, so you can learn more about different television technology to help make a better buying decision. And our product and technology — so we have created this brand new search engine using proprietary ranking algorithms. So we have 3.2 billion pages of documents that are only shopping; that compares to Google’s latest published numbers of 8 billion pages of everything. So we’re just specializing in all the sites that are shopping oriented. And — our error technology is sort of our core crown jewel of our technology. This is a patent pending technology that looks at not only the incoming links to a page; but also outgoing links, and we also look at the context of where those links are coming from. So a bunch of links are coming from off-topic pages or spam pages, we don’t give any additional ranking for that; but if it comes from other sites that are related to your topic, product, or shopping, then we give you a lot of ranking boost. So it’s much more targeted and relevant search result and much more spam resistant. And we integrate that — Sure, I’ll make it real quick. So, it’s the only site that combines the research and shopping. Our business model is CPC; we have over 5,000 merchants who are selling their products on our site; over 20 million products; and our revenues are ramping up rapidly. And founded by founders mySimon and WiseNet, and management team from Yahoo, eBay, Alta Vista, and Nextec; located here in Mountain View. We have launched a Japanese company joint venture, who have 42 employees here, and 18 in Japan, and we are one of the fastest growing shopping search engines. And, a brief explanation about our technology and kind of coverage that we got in the press. So that is all I have for right now. — [Applause] Gautam did you want to explain about Simply Hired? And just a quick question, between how many people have actually come to to search for products? What about Simply Hired? I think in Silicon Valley, it might be slightly bit more with all the job seekers. Oh no, it’s generally popular. — [Laughter] How many have accessed – oh there you go. So Gautam, did you want to do – Yes, first of all, thank you very much; it’s a pleasure to be here. I am Gautam Godhwani; I am the CEO of Simply Hired. And since Michael started off with a caution, I am going to indulge in the same So I am going to ask you guys to raise your hand, and keep it up, if you have ever looked for a job? Can I see a show of hands? — [Laughter] Hey keep them up, I want to see; I want to see the hands. — [Laughter] Okay, now, I want you guys to keep your hand up if you enjoyed the process? –[Laughter] So that’s kind of a value proposition for Simply Hired in a nutshell. Simply Hired, as you know, operates a search engine for jobs. We index jobs from a few thousand sources around the internet, including job boards, newspaper sites, non-profit sites, Government sites, and employer sites; anywhere where the jobs are, we go out and get them. Today our index is about 5 million jobs total, which if you know about sites like Monster and CareerBuilder, we are about 15 times the size. And when we started in March of 2005, to give you a sense of our growth, we were about 8 times the size. In addition to that, Simply Hired has a number of other features; for example, we integrate referral data, salary data, mapping data, so that you have a really, really great job search experience when you come to the site. The company was founded in October 2003; we came to market in March of 2005, and the company has now raised about 17.7 million in funding total. A couple of months ago, we announced partnership with News Corporation and launched MySpace, so we now power MySpace Careers. So if you have a chance, check out I think the excitement for us in the company really stems from the fact that we really believe we are working on an important problem. If you look at the life of an individual, you know, if you look at our lives, we believe that there is really three important decisions that each of us are going to make. The first is the spouse we choose. The second is the house we choose. And the third is the job we choose. And, you know, if things go smoothly for you, you’re probably not making those first two decisions very often, — [Laughter] — so that leaves jobs And if the Bureau of Labor and Statistics is correct, then we are each going to make that decision, ten times in the next twenty years. So that makes job search the most important decision that we will make again and again in our lives. And so Simply Hired is determined to help people find jobs they love and I would say that we’ve probably executed about 15% to 20% of our vision, so we’re pretty excited. There’s some great times ahead; so I am excited to be with you. Thank you. I guess there could still be a dating search kind of a site for the first objective — [Laughter]. There could, but yes, thank you. — [Laughter] Suranga, do you want to explain about Blinkx. Yeah, yeah, I don’t have a question for you; instead I have a request. When she asked, when somebody asked earlier, not in the interview to put your hand up saying you had been to So the first demand I have of you, well all of you, is to go to Blinkx tonight and have a look at it. — [Laughter] It is much more interesting, it is much more interesting than sitting here staring at me or listening to me, so. Blinkx is a video search engine; the company is about two and a half years old. We looked at, you know, my team — my founding team and I — all come from a sort of background in audio and video search and that kind of area. And when we first looked at the space about two and a half, almost three years ago now, we were amazed at how little technology was being used in that particular area. At the time, if you tried to look for a video on line, even for the few videos that were online, it was amazingly difficult to find things. You basically stumbled on things if you were lucky on a sort of an existing search engine, Google or Yahoo, that kind of thing. So we focus very heavily on changing that problem, of fixing that problem. We use all kinds of interesting technology, which we can go into later, using things like speech recognition, visual analysis, face recognition, all that kind of stuff to really get computers to understand what is going on inside video content. And we do that across the web, so, you know, I am sure all of you here have read articles or checked out video websites in the last six months; it has become a very hot topic. But what you will notice with a lot of those sites, almost all of them, is that it is all about stuff that just exists on their own site. Blinkx is very different; we’re about what’s out there across all these different sites. So from our single search box you should be able to find everything from a Bloomberg report on Condoleezza Rice’s latest flight to the Middle East, all the way through to probably a more compelling video, of a skating boarding dog. So, — [Laughter] — that’s just to sort of give you a few milestones the company – we’re still relatively small in the sense that we, you know, we’re sort of angel funded and sort of personally backed right now. But, despite that, on that shoestring budget, we have managed to get a pretty popular site and recently have started announcing a couple of pretty large distribution deals that we have signed. The first of which went out a couple of weeks ago; we now power Video Search on an AOL sub-portal called Study Buddies Education Portal, which is a very exciting sort of piece of news for us at the moment. The, from the point of view of kind of where we see things going, we are amazed that even today, video online is, if you look at what people do on there, it is pretty naive and pretty short sighted, you know. It is still sort of short clips that okay are fun for a while, but people don’t seem to grasp just what is possible, you know. The Internet is an amazingly powerful distribution network; it is very, very fast; it’s already piped into millions and millions of homes across the world. It is only getting faster and more pervasive; it also is very, very quick to adapt to new display technologies and devices whether it’s phones, whether it’s your TV, whatever; and at the end of the day, the media of choice for most of us is television or video content, and more of us watch TV than when we read books or newspapers or magazines. Whether you like that or not, that’s because of the motive nature of that particular medium. We think that intersection is really, really exciting; you know, it’s about technology, but it’s also about what a lot of people really like doing for lots and lots of hours of every day, and we’re hoping to play a small part in hopefully connecting those two together. Well, thank you, so let’s just hear some of the questions that I had from this panel and we will definitely open it up for questions and answers from the audience. But here is my question, and this is probably extremely basic to start off with, but it’s a question that usually gets a lot of interesting responses. What is vertical search? Here is the reason I ask this question because my mom, and I know it’s sort of a cliché person or term to use, — sorry, person to use, when asked her what’s a vertical search site, she says it’s Expedia because she could be able to look and search for information about travel deals and so forth better. She thinks could be searched as well, but there is definitely a difference there, and what you guys do is just spider out and be able to collect that information and pull it and sort of get people to sub-web that. So, my question to you is 1) what do you think is vertical search, and 2) how important do you think it is to distinguish yourselves from the sites, which just sort of takes submissions? You are actually going out there to gather information from the web. So Mike could you just start off this? Right, — so I think that vertical search engine is different than general search engines, like Google or Yahoo, in that general search engine gives you all kinds of documents that match the keyword that you are looking for. So, vertical search engine is really a specialized search engine that has a separate index or database of information that is specialized in the one particular topic. So that, you don’t get a lot of these off topic or unrelated irrelevant results to save the user a lot of time in getting to the information that they’re looking for. So whether it’s shopping, medical, job, video, these are all vertical search engines; so there is clearly a need and a room for a specialized search engine of all different kinds and different baselines. Gautam? I guess I would say that by definition I think the reason that search as a whole emerges is because you have something to search, and if you have just a few items, search is trivial. You don’t need to search; you can just see it. So I think general search emerged because there was a lot of web documents out there, and as the number of online destinations has proliferated, I think you see that now there’s a lot of documents or a lot of information within a lot of specialties. So let me give you the example of job search. Today, in addition to the big boards that you hear about, Monster, CareerBuilder, there are 5,000 other job boards; then there’s 3,000 newspapers; there is 5 million employers more and more of which are putting their job sites online, and by the way — every state and county has their own job site. Now I’m going to tell you that the biggest job site out there, like Monster, has only 10% of the jobs that you might be interested in. So what are you going to do? And that is the reason that the vertical search emerged; but I think that the long-term promise of vertical search actually, I think, is going to make that term a bit of a misnomer. Because I think the vertical search players at the end of the day are specialists. They are specialists in a particular area that requires great specialty, and when you look at job search and what a complex process it is when you go thru this three to six month job search process, where you do everything from go and search for a job, filter companies, see benefits data, find referrals, create a resume; that’s where the specialty comes in. So I think that you are seeing an era of specialists emerge, and that’s fundamentally, I think, the most exciting part about vertical job search. The fact is that Simply Hired is known as a vertical search company, but we are actually an employment company. We help people find jobs; that’s what we do, and search is just one mechanism to do it. Is it important for you to make that distinction to people who come to Simply Hired; for example, to be able to tell them that we go out and get as many jobs as we can that is applicable to you, versus telling them this is just a job site, you should be coming here to look for, — is it important for you to point out that you go out and spider and get the content that they will actually need? Absolutely, and we go out of our way to point that out. If you go to our homepage, we’ll tell you we index 5 million jobs from literally thousands of different sources, it’s right there. Okay, good; Suranga? Yeah, I think there are sort of two or three hallmarks that typically yield a vertical search engine. You know, the first is essentially a topic or an area or activity that is of great importance to people at certain points in their lives, you know, and all these — the list you searched out earlier, all sort of fall in that category. But then it’s also about that content being out there and being, you know, there being lots of it, but it being in certain areas or certain spaces or certain sites. So again, the big problem with general search engines is that when you do your product search, you get dumped with all sorts of things that aren’t to do with products that still match the keyword for example. And then the final area, which is probably the most important, really, I think in the long term, is that this area that we’re talking about has to be one, which requires some kind of very specialized approach. Whether it’s to do with something as simple as how the interface interacts with the user, or whether it’s something more complex, like some deep technology requirement to actually understand the content that is there, the data that is there. That’s where it really matters, just to sort of not be, to sort of self advertise constantly; I’ll use an example like Zelo, which is a real estate search engine. What’s really neat about Zelo is not that it taps into any information that wasn’t already there; it’s all already there, and people who are really good at searching could already get to most of what Zelo could do. But what they really did that was smart was to pull all those things together. So you can type in a zip code or address or whatever and get not just, you know, the thing you are looking at, but also the prices historically, and all kinds of legal information, and information about the neighborhood, and so on. That’s really where it gets very smart, and to do that on a general search engine is extremely difficult. Even if you look at Google or Yahoo; yes, they’re general search engines; that is the core business for both of those companies, but they’re also in many ways, vertical search engines; they all have separate tabs. And so even in their case you can see there is admission of the fact that it makes sense to separate Frugal from Google web search or whatever else. I think, and also I mean, that’s sort of one of the reasons as well, right? The one search box is not something that’s really working anymore, and it has been great for so long that you could go to one search box, search for whatever you want; but then you want that kind of flexibility in what you want to search for. So keeping that in mind, and you can probably start to understand, what are some of the three main priorities that you have when you design the interface of your own search sites? There are sort of the search bar and you can click on search and that is what Google does; but when you designed your own interfaces, what are some of the three top things that you have to keep in mind or you try to keep in mind to make it easier for these people that are coming to your site? Sure, far from the lesson that’s okay get a good sense that, you know, while we are a video search company, absolutely in deed, in a sort of higher level what we are is more than that. One of the best analogies is of remote control that lets you access video content live. There is a bit like a TV remote control that lets you click through channels, we should be sort of remote control. And in that sense, clearly we’re a content remote control trying to do different things, so that’s the way we think of it. We literally go to a place of thinking that the average user, well all the different average users as they were, what do they all come to the system or site to do. And you will see that our site in that way fits two gloves. There is a search option because there is a certain section of people who are after specific piece of video or a certain piece of information they know what it is, and they are willing to type that in, and if they find it well, and click Go. Then there is also a whole bunch of people, who don’t really know; they just want to browse some entertainment or browse some top news story of the day or whatever. We cater to them as well. So I think, you know, search, even if it’s endeavors to become verbally simplistic, search has gone after the group of defining itself purely as search box with a button next to it; but it’s actually gotten more complex than that, and that is an important thing to bear in mind. We still use the search index, logically it is all about search, but you can have a very, very different — without search engine. Another good example is on Become, the fact that they have this concept of two bookings because when we are buying something, research and time are two very separate things. It helps you find information, how you access that technology can change quite dramatically based on what you are actually doing with it. So, in terms of our interface, certainly with a name like Simply Hired, I feel compelled to say that simplicity is the very first thing that we pay attention too. And I think the effort we’ve made is to really look at what the user really wants to accomplish at the end, and to be able to use a very directed interface to get them to that. So obviously when you go to our home page you’ll see that, in fact, it’s a very, very simple interface, very Googlesk if you will, because we believe that the user is coming to our site to accomplish a very specific task. Efficiency of the interface is the other aspect of this that is very important to us. When you have a database of jobs, for example, that’s on — having them find an efficient way to get down to something they really care about is important. So, for example, our filtering interface I would venture to say is the deepest on the web. You can search jobs by the size of the company, the number of years of experience on the job. Lately you can search it by whether the company’s dog friendly; so there are all sorts of different ways you can drill down literally with a few clicks, and that efficiency is very, very critical to us. The third area I would highlight is contextually integrated, and what I mean by that is that if you are in the middle of actually searching for jobs and you want other relevant information, you ought to be able to find it. That’s why next to every company name, there is a link where you can see background company information with financials and a link to the site. Next to every location, you can drop down a map and you can actually then start to map it and see commute distances. Next to every listing, you can see referral data, so you can see if you know anyone at that company. Next to every occupational category, there is salary data, so you can see how much the job pays. So what you have is a job search process that is very simple, very efficient, and highly integrated, because at the end of the day, if you have more information it just shouldn’t take more time, and that’s where the challenge comes in for us. I had a question about the dog friendly part, just a — She always gives me more questions, why. — [Laughter] So you essentially what you do is go to the company websites, their job listings page. Is that something that exists on those pages or is that knowledge that you have on these companies on dog friendliness? No, it’s knowledge that we have. A lot of the additional work that we’ve done is to essentially go out and figure out the data sets, the data stores that are relevant to our users, and to go out and get them and integrate them into the sites. So, for example, we actually in this case, worked literally with Dogster. It’s a social network for dog owners to say what are the most dog friendly companies, and they didn’t have it either, so we worked on that together and found a way to get that info. So whether it is salary data or company data or benefits data or anything like that, we will go ahead and do that. So to give you a couple of other examples, we’ve partnered with Care2 to highlight socially responsible companies. We have partnered with Working Mother Magazine to create a set of filters that are mom friendly or family friendly companies. And so that is really the focus of the company is to provide very, very deep targeting so that you aren’t spending more time, but less time. So there’s hardly knowledge out there beyond what you are able to accumulate, is that correct at this time? That’s correct. So, at Become there are three considered design goals we use – simple, clean, and easy to use. It’s pretty fundamental; keep it clean; keep it simple; and keep it easy to use. And we conducted a focus group study of our site in New York and LA a couple of months ago, and we were very happy to hear that they really liked the site because it was simple, clean, and easy to use. And a lot of companies put a lot of blinking ads and others, and I found out that consumers really get turned off by a lot of ads and a lot of busy things going on. They just want simple — a site that gives them the information that they are looking for. Now in terms of sort of approach or the philosophy, we use both structured data and unstructured data. And what I mean by that is Web pages are very unstructured codes or documents, and then there is structured information that merchants have. They have product images, prices, and descriptions and other specs. So we’re not biased toward anyone or the other; we just look at what those users need and what is the best information that we could present. If it is unstructured, we have to develop a crawler-based search engine technology to crawl billions of pages of documents. If it is structured data that merchants have, then we take that and display it rather than trying to extract that information through technology, because in those instances, technology can never be as good as structured information that is already available. So why do you try to do something that is not as efficient or better than what’s already available. So we combine, both structured and unstructured data in our site to make it more easy for people to find information. The last point is that we try to have our user spend least amount of time on our site, and that of sounds kind of counterintuitive, but we’re not driven by page views or CPMs. We’re driven more by how quickly can we help people find the information that they are looking for and just have them go on their way, whether it’s clicking out to another website or clicking out to go to a merchant to look at the product there. Because our business model is driven by cost per clicks (CPC) rather than page views or CPMs, so unlike some of the first generation portals that are trying to have a walled garden of information that they want you to stay there as long as possible, click as many pages possible, because ultimately their revenue model is based on the CPM. But our model is very different, we want people to spend the least amount of time on our site. We want them to come back often, but spend the least amount of time to just click out the information that they are looking for and that’s how we sort of measure how we – how well we are doing as a company and how the consumers find information that they are looking for. Is there a question? So during the research phase, does that generate revenue or is it only when they go to the buyer? Yeah, research, we don’t make any money; so that’s organic results, free objective, unbiased, but when people click to go to a merchant, by clicking on buy button or click on one of the sponsor links at the bottom, then we generate revenue. Actually that takes me to the next question that I had, thank you. [Laughter] The revenue model, and you explained a bit about it, but could you go a little bit more deeper into what is currently in place and how you’re planning to scale, because I guess there are two ways that comes out for examples of revenue predictions. Can you explain about that? We generate revenue through some 5,000 merchants who advertise on our site. They give us product feed of about 20 million products, so whenever a user clicks to their site to look at the product or purchase the product, then we get a small cost per click fee; and we have a standard rate card. We also generate revenue through Google AdSense, so we’re one of the larger Google AdSense partners. We put Google sponsor links at the bottom, and when people find some relevant merchants or other sponsors that has the information or product they are looking for, we do a rev-share with Google, so it’s all CPC. Okay. We have one CPA which is sort of a portion of the revenue, but that’s just one out of 5,000 merchants and it’s a very unique case there. But that would be sort of what most merchants might be possibly interested in as well, trying to push the performance of the — That’s right, but CPC is a better model for us and we think it’s a fair model because it distributes the responsibility, the risk, and the reward to both us and the merchant; whereas, CPA puts more burden on us rather than the merchant to try and convert that into a customer. Right. Gautam? My sense is that companies like ours, this whole vertical search category, fundamentally are lead generation companies. In our case what happens is that you have a candidate show up to our site, over the three to six month job search process, we develop a very deep profile of that candidate, and then we end up matching that candidate to targeted destinations. Now, whether that candidate is then delivered to those destinations via a click, or as a profile or a resume, whether they’re delivered to a job board or whether they’re delivered to an employer, we’re agnostic, because we make money in all those cases. So the way that the company makes money is using clicks. There are paid job listings where a candidate can click on a job listing that’s at the top of the page and go through, and we make money that way. And when a candidate creates a resume, we match that resume with targeted destinations that also pay us money. We don’t advertise on our own site today, primarily because we use our own site as sort of a space or an area to do some experimentation regarding such things as interface and how the system works. What we do instead is advertise through all of our distribution partners, so when a partner of ours uses the technology on their own site, we usually advertise and it’s placed along side the content. The interesting thing about video is that it rears the possibility of video advertising, which right now is extremely popular and sort of — both with users and also advertisers. But the beauty of our system is that it works perfectly well with text sites as well, and many of our partners use that instead of this stage, just because the market there is more efficient, more fluid. The area that we have coming up, which is very interesting, is the ability to sort of splice together content and advertising based on searches. So the interesting thing that we get is we know what someone is interested inn any given point in time; we know why they’re watching that particular video about Britney Spears or this particular video about, you know, the news or Tipper Company or whatever. And that gives us the ability to match that piece of content with a piece of very relevant advertising. If you look at video content online today, most of it does have advertising before or after or during it; but it tends to be very sort of, you know, dare I say bland, brand-oriented advertising, which works to an extent; but doesn’t get as far as it could do. We have the ability to sort of, I guess, bring the sort of ad words “esq model” of advertising into TV advertising. So we are doing a few trials in that area right now. Today there isn’t enough of that advertising to make it sort of scaleable on a Google or Yahoo, but that is definitely the future for that kind of content, I think. Actually giving, I mean, considering it’s video, there’s definitely a question. Is there an editorial sort of layer to; obviously just like Gautam was talking about, there’s sort of a second layer of how you go about filtering that information, but is there an editorial aspect to that and what do you consider to be sort of okay to put on your site, and what do you consider not to be that okay. Is there an editorial layer after you collect the data? Yes, so not on our site. You know, we try to spider everything, whatever it is, wherever it is, from the very satiric to the very main stream Many of our partners focus on a particular area; you know if someone is building a vertical portal on a particular topic, they might only have video from certain sources. We obviously offer a lot of filtering opinions to the users, so as a user you can select if you want things that are very recent, if you are sort of a news-oriented person, you can select what source you want your information to come from. Say you want information from only entertainment or you only want information from the sort of user-generated site of the video content world. You can also do things like, more sort of rudimentary filters, things like adult filters, and so on as well. We, — technologically it is actually quite a challenge, you know. A lot of the technologies that have been used to make Web search good and relevant, things like linking analysis, things like domain analysis, don’t really work so well on the video. Pieces of video aren’t really linked to each other in the same way that a Web page is linked to each other or blog links to each other. So, we have had to build a whole load of new technologies around how that works; but beyond that, that is all about relevance, it’s not about filtering. Filtering – the options are there, but we either let the user do it or the partner do it. Okay, that’s all. And considering we’re in Google, this sort of brings up the next question, and do what we do, there’s definitely out there, I am guessing Google job seeker came along or Google shopping, which is sort of Frugal in a way; but I’m sure you think about it a lot. I’m sure it’s sort of question on how do we approach this if Google got into our space, and if they did, how are you planning to sort of work around that? What is your sense of how much Google can provide and how you are competing with Google, and especially Google and Yahoo I would think. What is your sense of the competition from the general search engines? Suranga? Yeah, sure. Well, that’s a tricky good question for us. We worry about it a lot to start off with. In reality it seems like we didn’t have anything to worry about, at least so far. Google Video isn’t a search engine; it has search engine built into it, but it only searches it’s own content. So that sort of solves a very different problem, it does that relatively well, but it doesn’t really go beyond Google. If you do a search on Google, you’re not going to find a video clip, you know. And yet, many of our geezers want to find video clips, so from that point of view, there is no real direct threat. And, in general, we have seen that the bigger players online, Google, AOL, Yahoo – even though they have the search capabilities at hand, in some cases, haven’t really deployed those or gone after those. For them, I think, going back to the point Michael made early; it’s more about building, not exactly a walled garden, but certainly an experience where you stay on the site. And there are advantages to that, and the beauty of Google Video for example; all the video is played on, so there is a lot of control that Google can have on the viewing experience. For example, whereas, the problem that any search engine has is that once you click off, we can’t really control how fast that video plays and so on. We don’t want to do that; we want to be a gateway. We obsess also about getting someone from the front page off the site as quickly as possible, you know, as many times as possible, sure; but as quickly as possible. So, from that point of view, we’re pretty unique in our space. There are a few smaller companies that have gone after the sort of video search area, in the main, they’ve either focused on, you know, specific sub-verticals, which we think misses the point, you know. If you look at what our users search for, they search for everything, and the same user will come back and search for very different things, even in the same day or week. And then other players also are basically retrofitting existing technologies to try to sell video search, which again we think misses the point. I mean, a lot of video search engines out there still depend entirely on text, so it’s all about computers going and reading words that are supposedly about the video, like a title or summary or metadata, and that gets you so far, but it’s all too easy to trick that kind of engine. You can say your video is about one thing, and it can be about something completely different, and we’re already seeing that kind of spam happen a lot on sites like YouTube and Weber, and so on; where if you can get lots of views then you get to be cool, but your content may not be a cool as you could get and so on. So, the great thing about technology like ours, which actually watches the video and cracks it open is that we have a better stab at getting – making sure we find something relevant and accurate. And I guess even YouTube is kind of having a problem now that they are trying to put some sponsored content video — they have the Paris Hilton video on their home page, and the comments to that are really worth reading. But it’s definitely a problem that even a YouTube seems to be having in trying to sort of potentially make money out of what they have in terms of content and trying to get sponsored folks to come and put their content on there some where. Yeah, I mean, — unfortunately, I can’t tell you who it’s with; but we are about to launch a distribution, a sort of partnership, with a very large portal, very big US based globally large portal, and you know, they’re very interested in the broad demographic. They don’t want to just advertise to just teenagers or college students; they want to advertise to lots of people. And they want to have sort of a professional traditional content along side which to advertise. That’s what Blinkx can offer because we do have content partners like, you know, Forbes or Reuters or whatever, you know, that sort of thing. And, yes, there is a demand for the sort of garage video, YouTubes of the world and so on, but that’s only one part of the, you know, it’s like saying that all of online video will be America’s Funniest Videos, home movies. It’s not going to be like that; there’s going to be more to it. And that’s why I like the idea of breadth. Yes, that is important, but so is everything else. Gautam, — Google keeps you up awake at night thinking about it? Night? — So, I certainly would say that again companies in our category spend some quality time from time to time talking about Google, and as a philosophy, I wouldn’t like to see us really underestimate any company, much less companies that have market caps that are larger than ours. — [Laughter] So, — as a result, I would say that, you know, certainly with respect to Google, as a generalist, being a generalist like Google, it is very tempting to think that a generalist can be a specialist in every area. And as it turns out, that’s not the case because by definition a generalist is actually not a specialist. And to give you a perspective on that, for those of you that have searched for a job; if your job search stopped with search when you entered the first few queries, if I asked you how far were you into your job search after you finished entering the first few queries into the search engine, I suspect a lot of you would tell me maybe 5%, if that. And that is really the advent of a lot of these specialized verticals, which is that when you go deep, when you get specialized, there is just a lot more. For the record, I would say that Google does have a Google Jobs, it’s called Google Base, and when you look at the jobs offering within Google Base, you will see that in fact, Google Base’s objective is not to create a comprehensive job offering, but, in fact, to capture data that is not already in the Google index. That’s why Google departed from it’s philosophy of going to other sites and for the first time became a primary collector of data, where you can actually put jobs into Google cause they wanted to collect additional data. Now as Google creates this data store, they’re going to have two choices. One is going to be to create a walled garden to bring up the term earlier, and that would go against every philosophy that Google has; or the other one would be to open up the data. And if they do that, then Simply Hired will index Google data just like it indexes the thousands of other sources of data around the Web. The other issue that I suspect that Google has to contend with is that what they want to do is to continue to increase their advertising revenue. If you saw that slide earlier, that’s really the revenue they want to increase, and when you face the fact that some of your largest advertisers are, in fact, job companies that are your competitors; you end up having a pretty serious channel conflict. And that’s why I suspect that Google’s objective in jobs has been very, very different. It’s to capture additional data, rather than go ahead and compete with its existing advertisers. Michael? So, I think we do sort of watch Google closely, and we do compete with Frugal on our comparison-shopping site, because there is definitely a similar functionality and a product search and comparison-shopping. But on our research side, where we find product reviews and buying guides, that is more of a specialized search for shopping, it remains to be seen whether Google will develop a specialized search for product information, like I personally don’t think Google will because Google’s goal is to try and get any and all searches to go through, and use the general search engines for everything, so it wants to be like a Swiss Army Knife, you know. Whether you want a screwdriver, or knife, or bottle opener, or wine opener, it wants to be everything to everybody. We beg to differ; we think that there is a need for a powered screwdriver if you want to screw lots of things to — lots of jobs to do. So, you know, it remains to be seen but we’re not terribly worried about Google getting into our space, but we compete more with Google general search engine because a lot of people are still going to Google for doing product research. And our challenge is to how do we deliver even more compelling user experience and better information quicker, so that user will start to migrate over and use Become’s vertical search engine. And we also get additional other traffic through people clicking on our favorites bookmarks, other websites that link to us as well as some partnerships that we’re working on. And then Gautam, where does your traffic come from? And I am sure partnerships are a big part of that. Most of — I’m sorry, go ahead. This is the final question for Michael. You mentioned earlier structured and unstructured data. Yes. So what you showed us was really, well the content you showed us was really organized, and it looks pretty well organized to be just proof of scrolling. So I guess it is data-fed from vendors and — is it data-fed from vendors? And if it is, isn’t it — how does this interfere or does it interfere with the search results? And isn’t the search results the results you are providing us by vendors who are providing or feeding data like into Google Base, for example? Are there two sources of data is essentially your question? Yes, that’s the question. It looks really nice, which it sure does. Right. So, I mean, you could try it out; but we have two buttons in the search box on the front page, you know, Research and Shop. So when you just type in a query and hit a return, on the left-hand side you get research results and on the right-hand side, you get shop results. The research results are unstructured data from our web crawl. We have 3 billion pages of web documents. And on the right-hand side are our structured data that we get from merchants, so merchants feed us data, all 5,000 of them, and there’s like 20 million — so it’s like pretty large volumes of data. And it includes product images, prices, and other — So how have you ranked them? So we have a separate ranking algorithm for both web search results, research engine; so it has link content, link analysis, contextual analysis; so it has got some advanced mathematical techniques. Speaking of the others, how do you rank – The other, we have a different ranking algorithm for those. So for that we looked at factors like the content. Is this the most relevant product to what the user is looking for? Is this offered at more than one merchant, so that it seems to be a more — And then we also look at the click through rate; is this product getting clicked more often than others? So we look at click through rate. We look at CPC bid amounts, so we have a dynamic bidding system where a merchant could bid higher just like Google Advert, and that’s one of the factors that we consider in ranking a particular product or a particular merchant in our comparison-shopping. So, it is two very different engines running simultaneously, and we present the data to the user in a way that is very clean and organized so that it makes it easy for a user to go over what they want to. But this shopping person, the person is actually purchasing something. The listings that they see are not the listings that you spider for. They are listings that you go out and get from other sites, but those aren’t the listings of people who have submitted them. So our product search results shows products from – submitted by the merchants. There is no data coming in from your spider company? That is correct. We provide a research bit length, where if you want to research that other web that we automatically show you – you can click on that and you get the research results that are from the web. Okay, — and there is no real mixing of data? We are not mixing the data. — Okay. — Yeah. So is it possible that you can find something that is on the research results, but you can’t buy it; it’s not available? Yes, yes, because, you know, on our web index, we have some 200,000 merchants, and out of the 100 million products or more; so it is very, very comprehensive, and sometimes, we don’t have the merchant — products from the merchants even with the 5,000 merchants that we work with, everybody from Amazon to Wal-Mart to Target. But a lot of times they are offered a small merchant somewhere. So sometimes our users find products offered through merchants on our web index rather than from the product search results, and that is part of the power of the comprehensiveness of the Web search. You need to go to AdSense. I’m sorry? You need to go to AdSense. Right, AdSense, which is another source, so we like provide information from the web, information from the merchants, and then the Google advertisers and between those three, 99% of the time, people find information on the product that they are looking for. Do you find that it’s necessary for you to share that information with the searchers? So, I guess, if I go about, would that be information that I would find on that page which specifies that when you research, it’s information that we go out and seek, but when you actually purchase, this is information that is submitted by the merchants. Is that something that you find it important to specify on your about page? I think that people have sort of certain expectations of search engine results as well as comparison-shopping results. So — and we offer many different sort of drill down capability by looking at particular brands. So, I think – it’s – we haven’t had a – it’s pretty self-explanatory, our search results as well as product search results. — Okay. So people go back and forth between the two pretty easily. We can always use improvement on how we make it more usable. So, I’d be happy to talk to somebody. Okay, there are a couple of more questions and then we will go ahead. Is there any linking between people who advertise on the paid results and indexing of the result that comes up on indexing? Oh, is there any linkage between the product merchants and the web results? No, completely – it’s like there’s so much distance; there is no linkage at all. The Web search engine is through our crawling and through our indexing and ranking. That is completely separate from our product search results that are based on the data feeds from the merchants. Would they be involved from somebody providing the data feed that is not crawled? It’s possible; it’s possible. We’ll just take one more question and then we were going to the Question and Answer session, if that’s okay. Michael I have another part of the question, are the product listings sponsored listings? Yes, they are products that are provided to us through the merchants that we have a relationship with. It would also have to be advertised and data fed? Yes, they do pay us on a CPC basis. If I — I don’t want to be part of the content then I cannot be in that? That’s correct. And this is actually important to the quality control because if we make it free and open to anybody, then a lot of spammers, they just submit all kinds of junk data, and this is what’s happening on Frugal. So the user experience is quite poor on a lot of the product searches on Frugal, because, again it’s free, and when you have free, people don’t care, they just want to get their clicks. So they would put wrong descriptions, just to get their click. They would use all kinds of spam techniques so that the user experience becomes – suffers as a result of that. And Google Base is another example; they just through in recipes and job listings and product results and you get all kinds of hodgepodge. So you need a certain amount of quality control and the best way to do that is by having an accountability system, where –. Well this could give you an opportunity where you can really develop a technological program to get the spam. — [Laughter] True, true, but we also need to generate revenue to continue the best, so it’s an important issue, but we think that our current CPC model is good for the user as well as for the merchants and us. And is that something that is popular across shopping search engines usually; for example, – do most of the shopping search engines work on a similar model? Yes, most comparison-shopping sites work on a CPC model. They have some crawled results and they don’t get paid – that’s just like us, we don’t get paid when people click on a crawled search results, even though it’s a merchant. Then, I guess we can go back to the questions about partnerships. Yes, the sources of traffic? Yes Most of our traffic actually is word of mouth, organic. We do receive some traffic from the search engines, but not significant, and our partner traffic has continued to ramp as the other significant source of traffic. We are a very, very partner focused company, particularly the relationship that we announced with the News Corporation and Fox Interactive Media; so you are seeing us on sites like MySpace and others. But still, our organic traffic tends to be the predominant source. Yeah, the same for us. We’ve sort of dabbled with search engine results but it’s not really a major thing at all. Again, we’re amongst the very powerful for us and also with true partners we are getting increasing amounts of traffic that way. And in guessing the results of a strong sort of PR focus in trying to get the word of mouth kick started and going about. Is there any suspect sort, I’m guessing blogs or any social sort of viral marketing aspect to how you are getting your name out there whether it’s Blinkx or Simply Hired, or Become? What are your strategies and how you are getting your site to be popular? Well, obviously, if I were raising funding right now, I’d say immediately that my site is completely viral; everything we do is viral. — [Laughter] If you don’t say that, we’ll get shown the door straight away. [Laughter] — It’s one of those meaningless terms in the Valley. But, there are a lot more aspects to what any site is – e-mail in particular is – is a – just the whole concept of word of mouth. The fact that someone uses the site and finds it useful, and then tells a friend they found it useful is in itself viral. So we’ve done various things on sites to make that easy, you know, whether it’s a search you like that you want to save or share. Whether it’s a result you find you want to save or share, you can do that sort of thing, very easily. That definitely works, but we found that, you know, however many viral mechanisms you build into the system, most people would actually share things in their own way anyway. So, you know, you could with some content on Blinkx that actually plays from Blinkx itself. You can embed those videos on your own site for example. You can tinker an e-mail to a friend and they’re all like, I’ve got an e-mail for you and so on. But actually what we found through some interesting analysis is that many people don’t do that, but they’ll just copy the URL, and drop that into an e-mail themselves and send that. So, I think that as long as you build a fairly compelling service, and, you know, stick to your guns on that, then you’ll find that people will sort of push it around. Beyond that, you know, blogs and so on, there is definitely an opportunity to offer more services to those sort of fast growing parts of the Web, whether it’s the social networks; whether it’s the blogs. And that is something we will look into more in the future. But, right now, we’re getting pretty good growth just from straightforward word of mouth. So is there a question then? Is there a question? How many hours of video do you archive and how fast? Yeah, so — index rather than archive — there’s a tiny bit difference, but right now I think we’ve got a sign that says 4 million hours. I can, exclusive for this conference, I can tell you it is more like about 6 million hours to date. The speed depends on how we analyze the content; some content is analyzed more than other content. One very obvious distinction I can give you as an example is that Podcasts have no visual analysis, whereas, video content does. But even on the most analysis heavy content, we can do it in about twice real time, the speed of the content itself. Now some content can’t be played at twice real time, so that doesn’t help you; but that’s how fast we can access things. So, for example, if there’s a news story breaking and it appears on and BBC and go to UK, and a few of the big news sites, we’ll have it within, you know, if it’s a one minute or two minute click, which it typically is, we’ll have it about two or three minutes later at most. And that accounts for all the different things that can happen. The speed at which the index is growing is difficult for me to say, it was sort of doubling on a six-month basis, but it seems to be accelerating, if anything right now; but, of course, a lot of that is content, which is very long tail in style, which probably isn’t going to get looked at. So, I don’t know how meaningful that number really is. But, yeah, as I said, right now we have about 6 million hours worth of content in the index, and it’s everything from skateboarding dogs to Condoleezza Rice. We have one more question for Suranga? Yes, I have one more question for Suranga, and it’s about indexing video search. Am I to understand that the state of the art of indexing the actual content of the video, that means like if I am looking for a specific catchy, same songs with that line? What is going to be available, you know, current development, you know, tagging content of streamed video and actually going straight to that segment? Um huh, yeah. That’s what works with YouTube I believe; it’s a tagging aspect that is going to help me pull up a particular video. Yes and no. So, it’s a complicated question for a number of reasons. The first reason is, you know, technology itself is so – the actual state of the art video search is amazing, very, very powerful stuff, and I have some demos if I had access I could have shown you that. If we have a demo which works for a large enclave of movies, for a particular movie studio, and this is a prototype I’m building with them, and you can literally type any line, any of those movies, and it will jump you straight to that point. It’s very, very cool, you know. You can think of – they have a big archive, and some classic movies that you will remember some really cool lines from; you can jump straight to it. It will show you every time that line is repeated, so imagine, you know, when something like say Star Wars, or you get a line repeated over and over again – you know, use the force, and you get to see every single ‘use the force’ throughout the whole movie. — [Laughter] I get the feeling this audience might like that. [Laughter] So yeah, it does that kind of thing – very, very powerful. And, you know, that ties in both – you can then overlay visual analysis. So, there is some actually pretty good these days, sort of facial recognition based stuff, particularly with movies and produced content because actually you may not know this; but lots of stars like to have lots of scenes where their face is head on so that they are getting brand recognition. So, Tom Cruise did that. There are parts in Mission Impossible where Tom will stare at the camera for a few seconds, — [Laughter] — and it’s true. And the technology is good enough to allow you to pick out that kind of take. So, you can say ‘use the force’ but I want to see it when this character says it or when this character’s face is on the screen and so on as so forth. So, that all exists technologically. There are a couple of problems porting that to the Web experience that you will see on a site like Blinkx today. The first problem is making that sort of broad and fast enough to look across all this data, and that’s really, really hard, so. While it’s easy to sort of pick a 100 faces and say we could swap all those 100 faces – it is very difficult to pick the 2,000 or so celebrities that you would actually need to be able to track the sub-content that’s searched most for on our site, because the scalability issue sort of hits you there. The other big issue that we hear all the time is content. So things either jump to a line in a peak, in a movie, or in an episode, you know, that’s only really compelling when you have so much content that becomes a problem. So for example, my favorite Simpson’s joke is the one about the psychedelic – the psychedelic chili, which you may or may not remember, Homer ate once; and I typed that and it jumped straight to that line and showed some of that particular scene. But today, you know, I guess it’s Fox who owns that content, hasn’t put all that cartoon online, so although I could fill your demo with a bunch of DVDs showing you that in action, I can’t actually make that of any use on Blinkx TV, So, for that reason, the state of the art is very, very powerful. It combines a lot of speech-oriented analysis, a lot of visual analysis, things like facial analysis. The reading of text onscreen, works extremely well with sports content because if you think about things like an American football game, you often have information about the current plays; you can jump straight to a particular part in a game, things like that. This is stuff that people would love to have, you know, if you are a baseball fan or a football fan or whatever, but you can’t apply lots of it because it either doesn’t scale technology or, you know, even more frustrating, the content isn’t yet online. So there’s no way to actually share it with people. I think the question was – about how to get the word of mouth going? Yes, sorry. That’s okay. Go right ahead. So in terms of actually getting word of mouth, I think we were talking about media coverage as well. Just after the site initially got going, we found that the organic traffic grew just based on word of mouth, but we were very fortunate that we received a fair amount of media coverage, and I think that happened because the category was generally being very well covered. Search has been hot for a pretty long time, and vertical search has been hot, so there was a lot of coverage within the category and the company was fortunate that some of it’s own features and some of it’s new launches were covered as well. We do — we did do some other things, for example, within job search; there’s a lot of interesting stories to tell and the company owns a site called Simply Fired. — [Laughter] I would tell you — I’ve seen that, it’s a really, really good part. It’s bitter sweet to say that that was our number one source of media coverage. — [Laughter] But anyway, I would go so far as to say that – yes, people like Simply Hired and they loved Simply Fired. — [Laughter] I have a question on meta data and vertical search. [Question] Michael – do you want to answer? Yes, so — if I understand your question correctly, this area of user-generated content is a very, very interesting area for us. And, so we are currently developing new social and community features for our site to allow users to not only rate and review merchants and products, but also create wish lists of all different kinds. And create any kind of liss you want to create based on the particular passion that you may have. So whether it’s motorcycles, or anything that people have a passion for, they can create a list of products and share that with other people who have a similar affinity. So, that is one area that we are actively developing right now to offer more community and social services to the users. I’m sorry, actually could we just shoot a couple of questions, and then we can open it up for a Question and Answer if that’s okay? I had one more question especially to Gautam, and I think it’s also because there has been so much news about how sites are not allowing their content to be spidered. They’re not allowing their content to be shared. And I think a huge example of that was Oodle, which was a classifieds search site and I believe Craig’s List was a site that did not allow them to be spider. Is that something that you would call often or rare? I guess not often, but is that something that is a bit of a challenge for you to handle? The answer to that for us is pretty simple. It hasn’t happened to us, and the reason for that is that people are coming to our site to search other content sources that they want to get to. And those other content sources, as it turns out, want people to get to them. So every time that somebody comes to our site, as you know, when they click through, they go to a place that really wants that traffic in general. Where that doesn’t tend to be true sometimes is with the really, really large sites that have big sources of traffic, but even there, we found that because their space is so competitive, it is possible to build really great relationships with them. And while those relationships today are not formal relationships, we continue to engage in a dialogue with pretty much sites at all levels. I am including sites like Craig’s List that we have actively reached out to and talked to multiple times. So from our perspective, I think what we want to do is to be able to help the job seeker, but also be able to work with a lot of the other sites out there that themselves are working very hard to make sure that they get jobs on to their sites. So there really isn’t, I think, a sense from us that we want in some way to keep those users there. I think that was the point that Michael made earlier, so as a result I think we have a very, very good relationship with a lot of those sites. It’s difficult for me to say, what will happen there. I can tell you, for example, today that an increasing percentage of our database and a very significant percentage of our database actually comes to us not via crawl but via feeds, and that is probably the most significant statement I can make in terms of how complimentary these sites see us to be. Suranga is there some kind of challenge in how you are getting content to your site? Yeah. No, not really, I actually would sort of echo exactly what I just heard in the sense that it’s difficult I think for people who have some kind of inherent child conflict, which is probably the challenge that — I’m guessing here. But it’s an issue that AOL has to deal with right now in the sense that they have a video search capability in our space, but at the same time, they have a lot of video content themselves because they’re part of Time Warner. So when you do a search, do you put the CNN clip ahead of the ‘somebody else’ clip, or the Fox News clip let’s say. And we’ve never tried to do that; we’re trying to be a gateway. We want to be a place that feeds you to somewhere else that watches the video. You know, playing video is a very expensive thing, the storage of video, the hosting of it, the boundary involved in it, it’s nontrivial from a fun friendly point of view, so we don’t want to do that too much if we can possibly help it. No, we’ll do some little things, but we would rather you went and saw it somewhere else, so because of that we have a pretty good relationship. I mean, we’ve actually you know, as well as sort of broad based spidering of the web, we have something like 80 or 100 explicit partnerships where people actually feed things to us on a regular basis to make sure their content gets into our system. So that’s pretty good sign for how happy people are to sort of — happy for us to spider their content. One of the things I was going to bring up was the previous question out there about Meta data, because it’s an area, which really frustrates me. Meta data is one of these things, which if it worked, would be amazing and create a utopia and we would all be happy and run around and — [Laughter] But it’s just amazing how badly – actually how badly is actually implemented. In our world the big Meta data standard that everyone is talking about this month is Media RSS, which has its faults. For a start, it’s attached to RSS that works for things that happen over time, but not for things that don’t. But putting aside that kind of argument I have with it, it is amazing how even the organizations that are heavily involved with it, for example, a large Web company that isn’t Google, that’s based near here, you know, that’s heavily involved in it, yet still won’t do everything in it for whatever reason. I’m not actually blaming them for being difficult or anything like that, it’s just the natural reality of the fact that it’s very difficult to define a standard that’s both shareable and effective, and well defined, yet works in all these different context. And that’s why we’ve seen that there really isn’t very much sanitization, so where there is we embrace it whenever we can because it makes our lives much, much easier, but the reality seems to be that it just doesn’t – there is no single method standard that works for everybody. The other big problem is that even if you do find something that is so broad based and simple that it could work for everybody, you immediately get, particularly with any kind of active vertical area like what we are, you get immediate sort of stab effects, which are very, very difficult to combat. So you know, it’s the same with Web pages, you know. If you remember the old web search engines, you know, they used to use just the meta tags. That was when they searched Web pages based on tags, you know. It was an amazing thing when Alta Vista decided to read the whole page, and people was like, why do we need to read the whole page. It became very obvious because otherwise people would just use the word sex over and over and over again most times and that would get the page up. So, — we did the same with our thing. It’s amazing how, you know, if there is something big in the news right now, people on YouTube will type their video with that term even though it’s got nothing to do with their video. So, for that sort of reason, we are always inherently a bit suspicious of Meta data, and that is why we will try to go to the data itself if we possibly can. I will just finish up with Michael and then we can open it up to Questions. So problems in how you’re getting content. Right, so, I echo what the other panelists have mentioned, we haven’t had any problem with crawling sites. In fact, all of them want more visitors to their website, so it hasn’t been an issue. We also obey the rule about the protocol, so if a particular website doesn’t want their site crawled, they can just put up a robots.txt text file, so we honor that. So, it hasn’t been a probably for us. Actually the part about RSS, is that something that it’s kind of the pull versus push, as to how you are trying to get your content to the people. Job Alerts, for example, or if somebody is able to subscribe to a particular feed of — coming from a particular source or a particular type, is that something that you offer on your sites or does it matter? Yeah, we do. I think to us, technology is like our accessor, really one step toward the fundamental fact that frankly, and I hate to say this within Google, but I don’t really enjoy searching information. The relevant information should just come to me, and I think that’s really the place where we all want to move to and ideally it should just come to me if by just knowing about me as opposed to me having to specify that. So, — [Laughter] — these technologies that are coming up, I think, are one step in that direction. Suranga? Yeah, I mean, we use ours for exactly that sort of feature. Any search can be done at RSS speed, and in addition we have a vertical suggestion technology, which will hopefully do the right thing at the right time and so on. But it is still early days; I think there is a lot more to go in that direction. I think it’s rather — it’s how you want search to work for you. There’s a concept of, if you can go to a huge mall and just go through aisles and aisles of information and be able to pull up the one you like, or you can go to a smaller store where you just look for information and point, and sort of filter it down and say, I want that in that color and that particular size and be able to get that. So, it’s definitely a question of how you want search to work for you. I know somebody had a question out there, but we can open it up for questions at this point. [Question] I will give you a slightly different philosophy. We believe that in order to solve the job search problem and fundamentally what our company is doing is literally walking into work every day making a list of job seeker pains and going down the list and solving them. That is pretty much what the business is, and we really believe that the business stands on four legs, search, tools, content, and community. Because we are a small company, we tend to have a singular focus and so we very much look like a search engine today. But as you can imagine, when it comes to tools and various applications, you will see us do a number of things in that area, particularly related to resumes. And as you go forward with content and community, you’ll see us create additional functionality. I mentioned earlier there are a number of interesting areas of content, whether it is salary data, or benefits data, or company data, or others that we have small forays into today. The community part is very important for us because first of all, job search is a very, very lonely process as I’m sure many of us know. [Laughter] And secondly, even today, half of all hiring is done via referrals. Simply Hired was actually the first, we got some coverage right when we launched, because we were the first to integrate with linked-in and really blend job search with social networking. And you’ll see us do a number of other things in that area, so we really believe that those are the four components. I didn’t specify the percentages, but I think you will see us have a very singular focus on one of those areas at a time until we get it to a point where we are happy with it, and then we’ll move on. Yes, so, we’re 100% focused on shopping search to become the best shopping search engine, and within that shopping search, we invest heavily in our core research engine. That’s Web based crawler search engine and we also invest heavily in continuing to improve our comparison-shopping or the product search engine. We’re also investing in this social community shopping thing that I mentioned earlier. So, I think these three are sort of core legs of our offering; that’s research, shop and social community as it relates to shopping. Yeah. So, again I guess it amounts to, you know, we’re very focused on our particular vertical — the three ways we see people access our system and use it are through search, through browsing and through suggestions. So, search is obviously, you know what you want, go to the box, you type for it and start looking for it. Browsing is, you have an idea, but you don’t really know what you want and you start clicking to things randomly, semi-randomly and get more focused. And then suggestion is where you don’t really know, but — hey, you should look at this. Right now because we, as it’s still pretty early, if I were forced to put a percentage on things, I would probably put it as 80, 10, 10. We’re more focused on search. But I think over time that will shift. In general, I think the Web is going to move more towards that suggestive type model. You know, I agree with the comment made earlier that it’s insane that we always have to, our Web experiences keep starting, you know, little boxes and ‘search’ button. It would be much better if some of it gets put pushed to — I look forward to suggestion becoming a bigger part of that. I have a question for Gautam of Simply Hire. I’m just curious; I haven’t been to your site, I will. But, one of the things that’s frustrating is often times we go to Monster and other sites, you click on a job and it takes you to that company site and all of a sudden and they want to make us fill out a big thing, and I mean, do you get around that or do you just do the same thing? In the sense that maybe you provide more focused information from other sites that they give to you and you have your own that you gather and — but do you get around that aspect at all? Sure, that’s a very good question because I think entering your information again and again is an incredibly frustrating process. From our perspective I guess the simple answer I would give you is if you’ll be patient with us, just wait and see. Simply Hire is going to make some pretty interesting announcements in that area. We believe that the sites that are out there today are very motivated to make sure that the application process is easier and it’s more automated. And Simply Hire is actually moving down the path of being able to do some of that. So, over time in an ideal situation you would actually be able to apply with a single click because both sides know enough about you that you don’t need to go any further. And I frankly don’t see why that’s not possible in the near term. [Question] I guess I’ll tell you that it’s on our list of things — [Laughter]. [Question] Sure, so your question really goes to how much of it is feeds versus via calling. While we don’t put forward the exact percentages, it actually depends a little bit on how you measure because there are secondary and tertiary feeds and so forth. I’ll tell you that we expect that before the end of the year the vast majority of our data is going to come from feeds. So it’s a very very big area for us. [Question] So, are you referring to — are you — [Question] No, they’re mixed together. Do you when you say feeds, do you are you making a definition based on our assessed technical feed or a part that is paid — Or whether it is even paid feeds. So, by the way, when I refer to feeds I’m not referring to paid feeds; that’s not the implicit assumption that I’m making at all. Or, the listings that we have are sorted by relevancy. So, there’s no concept today of a partner paying in there at all. I’m doing the same; we don’t have a partner paying system — No. So, feeds for us means that we have a relationship with the partner where we get the jobs via feeds as opposed to having a crawling site. It has nothing to do with paid feeds. I was using it in a different context there. [Question] Sure, I think that when I talk about that I would argue that things do come to you. Anybody that uses a My Page, for example, or anybody that, you know, is going ahead and getting an Xstream with the RSS has things come to them. The issue that comes in is how much do you have to tell that site, that destination, before things start coming to you. And how much do you have to tell them before meaningful things start coming to you? And I think the challenge for these sites is to say that without you having to go in and do a lot of modification or putting in a lot of explicit data, the sites are actually smart enough to get you that information. And that’s really the challenge that’s ahead of us. In our space, we’re a bit fortunate because what you have is job seekers actually putting together a fair amount of information in things like the resume. So we have a bit of a head start if you will. But, again, we think we should be able to look at a user’s profile that already exists or be able to essentially glean information about the users that already exist and bring the relevant information to them. [Question] Well, LinkTen is one of our partners, so, once again, I guess, wait and see. I have a question for Mr. Yang. You talked about the community aspect that you are developing on your site where a consumer can rate a particular company based on their community support. Also these companies are your clientele in a sense. Is there any way wherein that community based ranking on their customer service will play into how you rank them on your site? In the sense, rewarding them for providing good customers by ranking them higher on your site that they may receive more place, that sort of thing.

, , , , , , , ,

Post navigation

Leave a Reply

Your email address will not be published. Required fields are marked *