As media companies search for new sources of revenue, everyone seems to be focusing on video as the new holy grail. Why? Because video advertising is one of the few remaining places in the media business that is still producing significant amounts of revenue. But how long can that last with the supply of video content continuing to increase exponentially?
Everywhere you look, you can see news and media organizations devoting more resources to video. Mashable recently restructured the entire company and laid off dozens of people so that it could focus more on video, and the New York Times has set up a team just to do Facebook Live video. BuzzFeed has an entire content arm called BuzzFeed Motion Pictures, not to mention several standalone Facebook pages that primarily do video.
The drive to produce more video is being driven by a number of converging factors. One is that traditional display advertising on the web is an increasingly terrible business—the amount of money that a traditional news or entertainment outlet can earn there is minuscule and getting smaller every day. That’s why so many media companies are driven to add pop-ups and viral-marketing “related story” links and other gimmicks.
Those gimmicks, and the data tracking that goes along with many of them, are in turn pushing substantial numbers of web users to install ad-blocking software, which further decreases the return on traditional web advertising. By comparison, the CPM or “cost per thousand impressions” that advertisers are still willing to pay for video advertising looks like gold because on average it is more than 10 times what a site gets for a print-display CPM.
The second major factor driving the mass adoption of video is Facebook. Over the past year or so, the giant social network has been focusing most of its efforts on video, to the point where CEO Mark Zuckerberg has said that video is the future of the Facebook experience. And wherever Facebook goes, that’s where the media tends to go, since so much of their traffic and audience reach comes from the social network.
Sign up for Data Sheet, Fortune‘s technology newsletter.
Facebook doesn’t have traditional in-video advertising like pre-roll yet. So far, it is only doing video ads or advertising adjacent to playing videos (revenue it started sharing with media companies last year). It has also been paying media outlets—including BuzzFeed, Vox and the New York Times—a rumored $250,000 apiece to experiment with its live-streaming feature, which helps explain why so many companies have Facebook Live teams.
Many companies are now bragging about their video views in much the same way they used to talk about web-page views, an almost completely useless metric from the old days of digital media. BuzzFeed boasts of having seven billion content views a month, much of which comes from video, and likes to talk about how stunts like the exploding watermelon pulled in 10 million views. But how much are those views actually worth?
As Gawker pointed out in a recent post looking at video metrics, the actual value of a so-called “view” is difficult, if not impossible, to measure. In part, that’s because what qualifies as a view differs from company to company. Facebook considers a video viewed if it is on-screen for more than three seconds—even if the audio is off and someone is just scrolling past. For YouTube, where the average CPM is $2, a view is anything over 20 seconds. So what is the actual value to an advertiser or user of that content? There’s no easy way to answer that question.
The problem isn’t just with Facebook. According to a study from Google last year, close to 50% of video ads on the web are never seen by an actual viewer, for a variety of reasons. That’s almost half of all the video advertising that appears online.
Snapchat has doubled its video views since last year. Watch:
When a media company compares the video views it is getting on its own site or on Facebook to the audience for something like an episode of The Big Bang Theory, the implication is that advertisers should be paying the same amount for both. But traditional television ratings, as flawed as they are (and they are flawed), can’t really be compared to the amount of time someone spends browsing a web page or scrolling through a news feed. As industry analyst Bernard Gershon told Digiday:
There are forms of video advertising that show some promise just as there are for non-video advertising. Native ads or sponsored content of the kind that BuzzFeed specializes in creating for brands can get significant amounts of engagement, and Snapchat claims that its ads are worth more because users opt-in to getting them by clicking on them, and they are more customized to the mobile experience than normal ads.
But those aren’t the kinds of advertising that most media companies are doing as they rush into video—most of them are selling traditional pre-roll, which virtually no one ever watches, and hoping that the price they’re getting for a three-second pre-roll ad on a 30-second-long video clip isn’t going to plummet the same way that display advertising on the web did.
Is that a safe bet to make? Probably not. Especially if Facebook starts to get serious about in-video advertising and turns on ads for the billions of video views it is already getting every day. That will mean another exponential increase in the amount of video and the amount of video advertising, and that means more pressure on ad prices for everyone. The problem with everyone jumping on a bandwagon is that eventually there is no more room.