We’ve been tracking ScanScout, a contextual video ad startup, since last year, and were interested to hear the company starting to open up about what it’s doing. Over the past few days we’ve been chatting with ScanScout CEO Waikit Lau to hear more about the interactive ads the firm plans to roll out this month. Here’s a screenshot and Lau’s explanation:
What you are seeing is a video ad that fired up on the screen after someone just clicked on the little “base” ad at the bottom. In this case, the video ad is a Public Service Announcement for Big Brother/Big Sister. You can see the original video content paused at the background.
Two-year-old ScanScout uses a combination of video analysis, audio analysis, and metadata to determine the context of a video and pair it with an ad (being especially careful about juxtaposing brands with sensitive material).
The company is making a pretty bold bet on its targeting skills, electing not to provide the standard CPM pre- and post-roll ads. Instead, 1) its system will be performance-based, meaning ScanScout gets paid only when people click though, and 2) it will not replace existing pre-roll or post-roll ads, but instead run a nondescript line of text below a video for viewers to click on.
Are people really going to click on ads around the periphery of videos? ScanScout currently places banner ads on sites such as Blip.tv, and Lau claimed that contextual relevance has driven click-throughs “extremely meaningfully much higher” than industry averages of less than one percent. The hope is this phenomenon will continue for higher-revenue video ads.
The new ads (as depicted in the screenshot above) will be shown by “dynamically inject[ing] a little script at the bottom of the video,” said Lau. “If you click on that strip, the video you’re watching will pause and then it telescopes up into an ad.”
ScanScout started with the problem of rich media search in 2005, said Lau, when he and former MIT classmate Steve Lee decided to start a company together. “We said let’s put in a grand each and buy a Dell server and see what happens—we weren’t going to quit our day jobs.”
Two months later, Lazy Sunday exploded on YouTube, and the two were blown away by Alexa traffic charts for the site. “So we quit our jobs,” recalled Lau. “In hindsight we weren’t really sure about a lot of the details. Worst case scenario, we find another job.”
In May of 2006 the team raised an over-subscribed angel round of $2.3 million. The company, which has 11 employees, is currently seeking additional funding. Lau’s exit strategy is quite transparent. “Advertising campaigns are most effective when the buy is across multiple media — there’s a lot to be said for a one-stop shop,” he said.
So if you’re so good at figuring out what a video’s about, we asked Lau, why stop with ads? Why not also do search? He laughed. “We actually have a search engine we built a year and a half ago which we never released. We’re not taking it off the table but B2C is very different than B2B. From a focus standpoint, it’s a very different business.”
Indeed, providing monetization services to video sites isn’t a bad place to be in right now. It’s helpful when your target customers all have their backs against the wall scrambling to make a living.