Startups can fix the Internet’s copyright woes, but nobody wants to pay for it.
The battle between Viacom and YouTube was supposed to be Sean Varah’s big break.
The media conglomerate in March filed a $1-billion lawsuit against the video-sharing site, claiming YouTube was dragging its feet by failing to curb the spread of unauthorized copyrighted content across its network. That same day, Mr. Varah released video filtering software that he was certain would solve everyone’s problems.
But in the months since, Mr. Varah, founder and CEO of MotionDSP, has had no takers; he has come to realize that what was once viewed as a huge potential market—video filtering—has become the technology that nobody wants to pay for. MotionDSP, and other filtering companies like it, are now scrambling to come up with different business models. “I could build this copyright thing, but so far, I haven’t seen anyone who’s going to write me a large enough check for it,” he says.
Mr. Varah’s San Mateo, California-based company, which had initially developed motion enhancement software to improve cell phone video quality, churned out its new filtering tools in just a few months in response to what looked to be huge demand for the technology.
But MotionDSP and others have found that video-hosting sites have little incentive to pay for something that isn’t going to reap returns, and media companies are reluctant to fund the filtering themselves. Investors are shying away, too. “It’s just a little early to figure out how people are going to make money from these tools,” says Neil Sequeira of General Catalyst Partners.
That’s bad news for the dozen or more companies that were hoping to cash in with their tools that can identify copyrighted video content and remove it from a web site.
Video filtering technology works by creating digital codes that act like a video’s fingerprint. But matching one video fingerprint to another can be cumbersome because companies must amass enormous databases of original content for comparison against material illicitly uploaded on the sites.
Once unauthorized material is detected, it is usually removed from the web site. That, says Ty Roberts, CTO of Los Gatos, California-based Gracenote, explains the lack of incentive for implementing the technology. “It’s really hard to get people to pay a lot of money for something that doesn’t help them make more money,” he says.
That realization is forcing companies like Gracenote, which also identifies music tracks for record companies, to seek out other revenue models. Gracenote is turning its basic video identification software into a recommendation system for movie and TV fans.
Even market leader Audible Magic, also based in Los Gatos, is barely making money from video filtering. The eight-year-old company has snapped up the lion’s share of contracts, but it has acknowledged that each deal is worth no more than $1 million annually—an amount that rivals say shows the market remains far too small. Audible Magic hopes to turn its fingerprinting business into a service that calculates and delivers royalty payments that video sites owe media companies.
But not everyone views video filtering as a dead-end street.
Redwood City, California-based Attributor, which charges media companies for crawling the web to identify copyrighted material, in December snagged $10 million from Sigma Partners, Selby Venture Partners, and other VCs. “We think that there are between 5,000 and 8,000 content owners who will pay for this, plus the top 10 video hosting sites, and then the 1,500 behind them,” says Attributor investor Marco DeMiroz of Selby Venture Partners. “There are several billions of dollars of opportunity in this.” Mr. Varah, and others like him, surely wouldn’t mind filtering out some of that.
RedHerring
