The world of online clips

Despite the broad popularity of online video, nobody's making any real money from it yet. It's telling that video search doesn't merit a direct link from the Google front page: the product isn't good enough, and doesn't bring in any cash. Google's video search has stagnated since the acquisition of YouTube, which began running in-video ads in May this year, but they aren't targeted on context as Google's text ads.

A couple of weeks ago, we discussed how content discovery technology could revolutionise the way information was accessed from the web. Last week, Blinkx confirmed that the future was already here as it beat Google, Yahoo! and MySpace to launch a system that "listens" to speech in clips in order to allow advertisers to target users more specifically.

The application that Blinkx launched last week represents a big test for the nascent content discovery technology, which has been talked up as The Next Big Thing since the internet industry crawled out of its self-inflicted dotcom pit in 2004. Blinkx's platform, which had been codenamed Project Trilby was launched under the AdHoc banner. It'll insert video ads that it deems appropriate at the start, in the middle, or at the end of video clips. At its launch however, the video hosts who allow the search engine to index their content, won't necessarily get a say on what product their users are being associated with, which could give rise to a few interesting debates in the near future.

It's likely that the main technical priority for YouTube has been implementing the copyright takedown software which, as a host, is vital for it to continue. Unlike rivals Metacafe and Revver, however, YouTube clips don't appear on Blinkx.

Another interesting fact is that Blinkx's which has head-quarters in the US and UK was founded by a Sri Lankan – Suranga Chandratillake. Even though Blinkx may have opened up a galaxy of opportunity to take context relevant advertising for video, it still has many challenges to overcome. Chief among them would be to find out whether "interruptive" TV-style advertising – i.e. not just a link or a banner on the side of the page – will ever succeed on the internet, because the success or failure of Blinkx will rest on whether users will accept potentially intrusive advertisements in online video. We should have some idea in less than a year from now, when Blinkx releases its first results as a public company. Listed on London's AIM in May, it achieved a market capitalisation of about $350m. The firm boasts its 111 patents have come at a R&D cost of $150m, and it shows: Blinkx's video search creams the competition right now, zeroing in on elusive clips.

Blinkx has also announced partnerships with RealPlayer and web search aggregator Dogpile, which is aimed at getting its search box onto more screens. It won't be long before this technology which is now used to place context relevant advertisements in videos, will be used to enhance the accuracy of multimedia searches and multimedia information retrieval systems on the Internet as well as on your desktop.