Due to technical difficulties, last night’s debate was held over until today.  Here it is in its entirety. 

The debate is on the topic of Video Search, and our participants are Mary Hodder of Dabble, and Gary Baker from ClipBlast!

1. Alternative Search Engine verticals seem to all have their unique “majors.”

General search engines have Google & Yahoo!
Job search engines have Monster.com
Business search engines have Business.com
and Video search engines have YouTube.

How would you describe your competitive advantage over YouTube? (Truveo?)

 ClipBlast! helps people search, browse, navigate and watch the entirety of the Video Web. ClipBlast! is the world’s largest evolving index of video content publishers and video clips accessible across the entire web.

Our competitive advantage over YouTube – which is not a video search engine but a video hosting platform — is our proprietary technology that automatically discovers video content publishers, learns about their video and metadata, and continuously updates the world’s largest index of video from across the web. ClipBlast! does not operated in a walled garden.

Neither does ClipBlast! host video; rather, we drive targeted viewers to video hosted by millions of video content publishers and platforms, thereby honoring their ability and right to earn revenue, exhibit their content however they wish, and build relationships with their audiences. This includes video from the major media companies, professional producers, news organizations (local, regional, national, global), new media publishers, vlogs, video podcasters, individuals, etc.)

YouTube helps people search and watch video hosted only on YouTube.com. Truveo and ClipBlast! both index the web for video. However, unlike Truveo – which is owned by Time Warner — ClipBlast! is independent and not affiliated with any major media company.

“General search engines have Google & Yahoo!”

This is not really correct compared to the rest of your list.  Google and Yahoo web search search the web, and as such don’t allow others to get their “data” (spidering) because it’s actually all out on the web and what they add to the picture is proprietary. So they want to keep it in house.

The three below allow their sites to be searched by others, because they are original content providers.  But they are the biggest in their respective spaces and so provide a kind of search, just because they have so much. But that said, they don’t have it all. and there is value in adding other more interesting niche items or bits of quality work to the mix.  In the end, I think the providers of content will be far more varied and distributed.. and therefore having search for all the stuff in a vertical can make a lot of sense.

Job search engines have Monster.com
Business search engines have Business.com
and Video search engines have YouTube.

How would you describe your competitive advantage over YouTube? (Truveo?)

Youtube is a hosting site.  It doesn’t have all the video, just a portion. Truveo is a search site, and while AOL does host video in other areas of the AOL property, Truveo does not.

That said, Youtube doesn’t have everything, for example they don’t have the Daily Show.  Comedy Central does. Dabble can search both and make them both findable in one place.

Truveo is pure search.  Dabble is human powered, so that we have a pure search platform, but layer human activity on top for discovery That human activity is very interesting, and can give context and relevance to searches in topic areas, in different times and places. And it is the basis for discovery which is 80% of the problem, verses search which is really only 20% of what users are trying to do with video.

Where would you position yourself with the other Alternative Search Engines (Dabble / ClipBlast!); PureVideo and blinkx?

Our positioning rests on three fundamental premises:

#1: Video Search Across the Entire Video Web

Where other Alts focus on aggregating content largely from user-generated hosting platforms like YouTube, Google Video, MySpace Video, Daily Motion, and a growing list of others, ClipBlast! is the video search engine that focuses on the entire Video Web. By that we mean our technology discovers, crawls and continuously builds the world’s largest index of video, from all publishers of video, as well as hosting platforms like those referenced above.

#2: Organizing the Video Web

This is the year the Web became the largest video distribution channel in history. There is now more video content available through your PC than any other device, distribution mechanism, or industry that preceded it. The Video Web has rapidly become the realization of Video on Demand – on steroids.ClipBlast! has been organizing and simplifying the Video Web for more than three years, so that people can instantly get to the most relevant video by searching, browsing and personalizing their experience with their PC. We focus our efforts on three constituent groups: (a) the viewer, (b) the video content publisher, and (c) the advertiser. This three-legged stool is the basis of everything we have done, are doing, and will do – especially given the ways in which video is triggering major shifts in how all three interact with the Video Web.

#3: Distributed Video Search Across the Web

With the deepest, most organized index of video available, our goal is to help viewers, content providers, advertisers, marketers and other search engines to get the most out of the Video Web. ClipBlast.com continues to grow as the leading Video Web portal, as users (now more accurately referred to as “viewers”) can rapidly search, navigate, watch and personalize their interface to the Video Web. In addition, we distribute tools and technologies built on the ClipBlast! platform, including widgets, custom players and our API, for greater reach and accessibility.

In the Search to Discovery continuum:  Truveo, Pure Video, Blinkx, etc are all algorithmically powered and require the user to know what they are looking for….WeShow, StumbleUpon and Digg are all human powered, and therefore have limited selection as made by the paid for or free editors. Dabble takes the best of search and the best of discovery, and puts them together in a scalable way.  No one else is doing that.

2. How does your search engine deal with the issue of Adult content?

 ClipBlast technology filters adult/explicit content. The Safe Search Filter defaults to the “on” position.

There is adult content on the web, and we index it, but put it behind the “unsafe” wall.  People can opt out, as they do in the real world.  But the idea is to let people decide how much they want to view and as Dabble adds more granularity this will become more powerful.  But for now, we just put the wall up, and as we semantically map the video, we push “unsafe” content to the other-side of the wall.

3. Future technology seems to be focused on looking “within” media.
E.g. words within an audio file (pluggd) and images within image files (PolarRose).

Video, of course, is both sight and sound - how is this type of technology impacting what you can offer your users?

 Speech-to-text and image classification are important to video search because they can (occasionally) add metadata that may not otherwise be available. Any technologies that improve the viewer’s ability to get to relevant video — or improve the content providers’ and advertisers’ ability to generate revenue and visibility — is important to us and is/will be included in our platform.

Yet it’s important to recognize that the world is just awakening to the Video Web. While viewers are becoming more comfortable with the infinite amount of video available through their PC (and eventually their portable devices), we must be careful not to put the cart before the horse by leading with technology.

The Everyzing CEO told the search panel audience at Video on the Web a few weeks ago that they are right with search to text about 93% of the time. *IF* they have one person clearly speaking, enunciating, with no ambient noise or music.  If there are other noises, or more than one person speaking, it drops to more like 70% accuracy.  And of course there is the problem that no one gives context in the speech of a video.  No one says, “hey, you there in the audience, this is a comedy!”   I think it’s useful info to have the speech-to-text and we are working on getting it to add to what we get now, but it’s only one part.  Like having a lengthy description but nothing else to search on. We want to use more data, but when the humans tell us what matters, we want to use that more strongly.  Right now, I think humans are stronger at context than machines.  That might change, but I have to build for what we have now, not something coming up later.

Also, regarding visual search, my friends at CMU and UCSD who’ve been working on this for 15 years say we are 5-10 years away from anything remotely simple and usable.  Think about scene cuts, and how when two scenes are placed together in one order, it can be comedy, and in the opposite order, it is a tragedy.  That’s AI to have a machine understand that.  And while I want it, I’m certainly not banking on that now.

“Video, of course, is both sight and sound - how is this type of
technology impacting what you can offer your users?”

Search and discovery and advertising can be quite a bit more colorful than text.  So we have thumbnails on the site that show a screenshot from the video, and ads will play that are more interesting.  And we have areas in our next version of Dabble were images (without sound) will play. It’s actually really fun to work on these kinds of things, verses pure text search engines. When I worked on pure text ones, all we had was text.  But this is really fun, having so much other stuff to play with.

4. What would be your unique goal for the next two years (2010)? Increased market share? #1 in your Vertical? Powering the video for other search engines? Acquiring or being acquired? Some combination of these?

 Yes to all.

Be the number one place to discover video, whether on our site or elsewhere, and have all video semantically mapped for great advertising matching.  I hate ads that aren’t relevant.  I think making ads relevant would be incredible compared to what we have now.  Without violating privacy of the watchers either for the search and discovery or the ads. 

“Powering the video for other search engines?  Acquiring or being
acquired?  Some combination of these?”

Powering video search and discovery is key.

5. Looking only at the other Alternative Search Engines, not just video search engines, but all of them, do you envision ways to collaborate with them, or do you plan to either go it alone, or to seek opportunities outside of the “Alts?”

It is our desire to work with Alts interested in the benefits of providing viewers a rich, Video Web experience. We love to collaborate with other search engines by providing our API and sharing our video index, as well as by discovering new and exciting ways to build an absolutely compelling overall viewer experience. That also applies to video content publishers who ask us to index, optimize and power video search within their own properties — which, of course, gains them the added benefit of being exposed across the ClipBlast! Video Search Network.

Absolutely, we’d like to partner, and have talked with a number of them.  There are lots of opportunities and since Dabble’s DNA is about partnering (we have 170+ partnerships) we’d like to in several different areas, not just hosting.  We are already working on a few now.

Many thanks to Mary and Gary. If you have a question or comment for them,
please leave it here, and we will ask them to check back occasionally.

This entry was posted on Wednesday, August 22nd, 2007 at 5:00 pm and is filed under Debates. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Sphere: Related Content