Blog Post

AT&T Gives Video Search a Try (You Can, Too)

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Video search engines haven’t found their place in the world quite yet, but after what happened with web search it’d be stupid to count them out. Now AT&T, after securing the exclusive license to use technology from the stealthy startup Divvio for the consumer market, is launching one of its own.

AT&T is calling its video search site VideoCrawler, and expects to open it up to the public in a week or so. For now, if you’d like to try out the site for yourself, AT&T provided a user name and password that NewTeeVee readers are free to use. Username: mediorite; password: beta0529.

Today we spoke with the executive director of business development at AT&T, Jim Stapleton, whose division is also developing the 3D web browser Pogo. He gave us an extensive demo of the site, focusing on the interface, which is what AT&T built.

First, the positives: I like the interface. AT&T built the equivalent of an AJAX start page for video, with modules for a search box, a player, search results list, playlists, etc., that can be moved around and minimized. That way users can organize the product to best fit whatever they want to do with it. The player itself is nice because it’s consistent, launching any type of embeddable media from within. I’m a little unclear on what’s included in the index, so as you guys play around with Videocrawler, please be sure to report back with what you turn up. For instance, I couldn’t find premium ABC content (the kind that launches in the Move player), but I could find Lost podcasts direct from the ABC site, which as far as I can tell aren’t usually embeddable off-site.

So far it isn’t evident to me that VideoCrawler’s results are markedly better than other video search engines. According to Divvio’s site, the company’s differentiating technology includes: the meta media player, continuous verification of broken links, ability to search for specific rich-media attributes, and easy addition of new sites to Divvio’s index. We don’t see any mention of speech recognition or other alternative search methods, but we’ll follow up with Divvio (whose CEO, Hossein Eslambolchi, used to be CTO and CIO of AT&T) to find out more about their approach.

Stapleton says VideoCrawler accesses more than 300 million pieces of content from 3,300 different sites, but in addition to video that also includes other rich media like slide shows, Internet radio and ringtones. For context, competitor Truveo says it has more than 100 million videos in its index.

VideoCrawler is making what I feel are some mistakes: It ranks videos by how many views they’ve had on its own site, not the larger web — a big blow to relevancy. When I brought this up to Stapleton, he responded by saying that, “We’re assuming that the user community of VideoCrawler will be representative of the broader Internet.” And instead of giving users direct URLs to share, it gives users a link to a VideoCrawler page with the video embedded. I worry that this speaks to a focus on building a destination rather than a tool for users. Stapleton’s defense: “This is a banner advertising-driven revenue model.”

I don’t see that VideoCrawler is different enough from the competition to really stand out. Apparently even Divvio felt the same way, because if it saw such a big opportunity for consumer video search why wouldn’t it have addressed that market itself? But I do like that we’re starting to come closer to a universal online video player. One way or another, let me know what you think.

8 Responses to “AT&T Gives Video Search a Try (You Can, Too)”

  1. I suppose I have a larger question. Why do we need to aggregate content into one master player environment? Do content creators really want to be subordinated to a larger brand like AT&T? I don’t think so and I don’t really see the value add in the long run. Viewers will find content that is relevant and means something to them. And they’ll want to play it in their own environment, perhaps a customizable browser where they can tag or bookmark their favorite “shows” (Oh we already have that – it’s called a web browser.) Then the show creators will come up with the player interface that suites their content and message. YouTube happened for reasons that will never repeat, and user-generated-content has had it’s day in the sun (it won’t disappear, but it’s not the end-game). Mark my words, this whole gambit to be a super “aggregator” – the next YouTube, will fail.

  2. Harlan Talbot

    This is not ready for beta. The app is buggy, the results don’t include relevant metadata (how long is a clip for example?, no descriptions), the advertising.com servers are slow…and when I tried to ‘log in’ with the user/pass you provided I was told it wasn’t valid. It’s not that I feel like I need to find a ‘video search engine’ it’s just that reading about it here give is some credibility, i tried it and even if I could remember the URL (remember WebCrawler…no…hmmm?) after this kind of user experience I don’t think I’d come back. Rather than let your technology drive your product development efforts, why not work with consumers to identify and unmet goal and define a product that serves that market instead of launching something that really doesn’t do much for anyone.

  3. We’ve moved away from a fancy search results page.

    1) We found that users just want the result, and be able to quickly refine the results>

    2) Once they do this, the “watch this page” link at the top of the search results gives them a full screen interface to watch the videos.

    3) While watching, there is a “link” link that enables the user to go to the original video’s source page and comment, share, etc.

    http://www.mefeedia.com/query.php?q=newteevee

    With this interface, the search utility is fast, watching is easy, and sharing, commenting, etc. can then be done at the source site.

  4. I agree about the results being in a weird place, and I don’t like how the page snaps you to the location. But I like the idea of taking an AJAX start-page approach where you could move them to your liking.

  5. editechial

    Very poor experience all the way around. Search results were not any better than if I simply used VEOH or even Youtube. The layout was also not executed well, does ATT really need to have $.25-$.50 CPM banner ads that clutter the user experience. Also, several of the videos I searched for came up as “file not available”

  6. At first glance, the UI for this doesn’t seem very good. The search is at the top of the page but the results appear in the bottom right. What’s up with that?