This week there is news that Yahoo! has joined the hunt for the video search engine, but like all those before it, not counting Google, which we have yet to see operate, as it is only just collecting content right now, it is a bit of a fudge.
On Yahoo!, the snippets of video that are indexed are all very short and all of them so far appear to be indexed, based on what the video is called or how it is described. The world is waiting for something that takes a view of raw video files and extracts a viable classification from them automatically.
If you key in the word football in the new Yahoo! engine, you are just as likely to get a film of just a football sitting there, a game of football, a football player being interviewed or the logo of a football club. And “football game” brings videos of table football and video games, as much as a video of a live football game.
The truth of the matter is that the reason that people have not yet brought out a video search engine is because first off it’s difficult technology, and secondly no-one is quite sure what problem it is that they are trying to solve.
Are we trying to find the name of a film, the date it was made, or a piece of action that the film contains, or the work of a particular actor or the first time Technicolor was used or short films edited with a particular software tool?
The answer is that a video search engine has to do all of the above, but more important than all of that is that there needs to be a business model.
As long as searching the web for freely available video is what both Yahoo! and Google are doing, then it’s not anything more than a few extra hits for their existing advertising customers. But what more could it be?
Well, for a start it could be a little more like a P2P network and when you ask for From Russia with Love, not give you endless trailers and films about Russia, but actually have some kind of index of key copyrighted acts and selected that old classic film.
One business model is that the search engine might become a front end for online copyrighted video archives that are for sale or rent. But if they’re going to offer this, then they must offer it at a different level from the hobbyist video fun that abounds on the internet. An advanced set of dialog boxes, defaulted for the trivial and free video on the internet, but tunable so that genuine films are searched for, can be easily set up. A simple question like “Are you looking to buy some entertainment” with description boxes, would get us started here.
They would also need to build a clean way of delivering the video directly to the consumer, rather that sending them back to the original web page where the video has been taken from and that would mean carrying out the transaction for film owners.
The reason that Yahoo! and Google see no reason to attempt to build something this subtle is because there is so little genuine film or TV content legitimately available on the web, but that is all set to change when this summer Sony makes its top 500 films available over the web.
Presumably all of these films will be encrypted and hidden behind password protection, so how would Google and Yahoo! deal with that? At the moment it is assumed that the existing models they are working on would need to index the trailers for such films that are resident on the Sony site and lead people there.
The weakness of such a model is that it still leaves all of those transactions in the hands of Sony, and doesn’t allow any other content maker to piggy back on the site, and put their own content up. Faultline is absolutely certain that once a generic search engine begins to transact film sales, then it will attract a huge following overnight, but until then it remains a chicken and egg story, and a question of which comes first, the films on the internet or the internet video search engine.
Yahoo! in the meantime is trying to promote the idea of video creators sending links of their content via Really Simple Syndication for inclusion in its database, which perhaps will improve its showing, while specialist video search companies such as Blinkx TV, are consigned to doing deals with tiny operations such as the one it did this week with London TV, a tourist TV service, rather than gets its teeth into genuinely desirable entertainment.
Copyright © 2004, Faultline
Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of the week's events in the world of digital media. Faultline is where media meets technology. Subscription details here.