Skip to Content

Vying to be Captain Video

With the launch this week of Truveo, the competition between providers of a new Web service–video search–is heating up.
September 16, 2005

Every tale of bootstrapping technology entrepreneurs seems to involve a garage. For Tim Tuttle, though, it was an apartment. “I actually emerged with a tan–from the monitor glow,” he says.  

Tuttle, founder and CEO of Truveo, a Web video search company that launched its beta site on Wednesday, spent two years in an apartment in Cambridge MA with his co-inventors to create the cutting-edge technology behind his company.

The video search arena that Tuttle entered when he emerged, just a couple months ago, is a lot more crowded than the one that existed when he went in.

Since May, Google, AOL, Yahoo, and Blinkx, as well as other sites, have added video search capabilities to their mix of search services.

As with text-based search, though, the quality and methods used by these services vary wildly. For instance, despite its dominance in text search, Google’s video search service produces somewhat haphazard results in response to some search terms, while Truveo seems closer to the mark.

Searching “Red Sox” on Truveo, for example, returned 388 files, the first 100 or so highly relevant to the Boston Red Sox–game highlights, features on the team, and so forth. Searching “Red Sox” on Google’s video search yielded 988 files, but the first page consisted entirely of clips from CSpan, weather reports, and other marginalia.

Search results vary because video search engines use different methods to hunt for video content. Most engines still focus on searching the text on a site, looking for obvious words, such as “video,” and for the file extensions most often found with video files, such as .avi, .mpeg, and .mov.

This method may be nearing the end of its usefulness, though, since many websites today are rendered on the fly, using a combination of applications and HTML. As a result, there’s not much useful data to be found by just searching the text for video information using video-search crawlers (software programs that comb through a site to assess its contents and report that data back to search engines). Many of the most popular sites with video content–CNN.com, for example–display no .avi or .mpeg links.

Tuttle says Truveo’s method circumvents this change in websites, by searching for application data instead of simple text. Truveo waits until the page is loaded before searching. This allows it to examine what is about to be displayed and to find video that text crawlers would miss.

“When our crawler looks at a web application or web page, it looks at it the same way a human would,” says Tuttle, who first began his work at MIT (undergrad, grad, and doctorate degrees) in 2003. “The crawler examines the visual characteristics of the page to find video and any metadata on the page.”

Like some other services, Truveo also pecks through a site’s metadata, contextual information supplied by the site’s owners or visitors about what can be found there. Most also look for any closed-captioning data provided by the video owner.

Blinkx software, on the other hand, actually plays any video or audio it encounters and pulls relevant data from the file itself. CEO Suranga Chandratillake says that for awhile, simply checking metadata and other page information worked well, since there wasn’t that much video around, and what was there was professionally done, with all the relevant information supplied. But now–with blogs, podcasts, and other forms of homegrown video, and an explosion of professional content  appearing online, it’s not enough to rely on these methods.

“Because it’s easy and cheap to create content and put it online today, things get messy,” Chandratillake says. “As a result, you need a pretty dynamic search technology.”

Whether or not Truveo’s or Blinkx’s technology will help it compete with giants like Google and Yahoo is an open question. But it’s clear that the market for online video is expanding fast, largely because broadband penetration continues to grow, making it more feasible for people to watch bandwidth-intensive video clips. Some 70 percent of U.S. homes with Internet access will have broadband by the end of 2005, up from 55 percent at the end of 2004, according to Nielsen NetRatings. And data from Jupiter Research indicate that, as of June 2005, roughly 20 percent of all Web users in the United States watch online video at least once a month.

What’s more, that percentage will likely grow as more major networks embrace the Internet as a vehicle for their content. CNN.com now offers free video (it used to be a pay service), as does MSNBC. Meanwhile, Comedy Central airs healthy chunks of its most popular program, The Daily Show, online for free. And WB will premiere a new series this fall exclusively on Yahoo. Clearly, after years of avoiding the Internet, the big media companies are quickly, if judiciously, deploying video online.

So as more video content comes online, the importance of high-quality video search will grow rapidly–just as Google’s explosion came with a massive increase in the volume of pages on the Web.

Could an Internet version of TV Guide be far off–a service pointing Web users to popular video content, both time-sensitive and “evergreen” (stored) programming?

“Absolutely yes. You’ll see applications and guides that allow users to find video on the Web in unique and novel ways,” says Tuttle. “And this isn’t Buck Rogers-type stuff. This is all going to happen in the next year.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.