Facebook's autoplay video ads can be a double-edged sword. By having the ads play automatically once they pop up in people's news feeds, brands may get more people watching their videos. But since Facebook charges the brands once an ad comes into view -- but not necessarily once it starts playing -- the advertiser may be paying for ads that people didn't actually watch.
To make sure advertisers don't feel like they're being short-changed, Facebook is testing a new option -- though the company doesn't think it's the best option for advertisers.
Under the test that's rolling out to Facebook's own and third-party firms' Facebook-equipped ad-buying tools on Tuesday, advertisers can still pay once their video ads come into view -- meaning on a cost per thousand impressions basis -- or they can choose to only pay once their ads have played for at least 10 seconds, or on a cost-per-view basis, as Facebook is referring to it.
"One of the requests that we've been hearing in the market for quite some time is a way for our marketers to be able to buy based on guaranteed view. ... It's really for those marketers who value video views as the best proxy for a business objective," said Facebook's VP of global marketing solutions Carolyn Everson.
While it may sound like the change is geared toward marketers like movie studios that simply want someone to watch their trailers -- as opposed to a retailer looking to push product -- it's actually a complaint that Ad Age has heard from various types of marketers who want to make sure the money they're spending to produce and run video ads isn't being wasted on people who don't actually watch the spots.
Facebook's autoplay video ads are considered a way to get more people to watch an ad because they don't have to decide whether to play it in the first place. However it can result in an ad playing -- and an advertiser paying -- before a person decides whether they want to watch enough of the video to be worthwhile for the brand. If someone scrolls through their news feed and stops on an autoplay video ad for half a second before scrolling past it, they may have no idea what brand was behind the ad or what product was being promoted, but the advertiser still has to pay for it.
Advertisers have the option to do some back-of-the-envelope math to determine the quality of the views they paid for. Facebook counts a view once a video has played for at least three seconds, so brands could measure how many of the ads they paid for were seen for at least a few seconds. And Facebook also tells brands how many views lasted for 50%, 75% and 100% of the video. But none of those measurements will put the money back in brands' pockets.
Until now, advertisers were able to get better upfront video-ad terms from some of Facebook's primary rivals that are also competing for the $7.77 billion that eMarketer projects that brands will spend this year on digital video ads in the U.S. "The competitive nature of the environment is forcing everyone to move faster," said 360i CEO Sarah Hofstetter.
While Snapchat has followed Facebook's example and charges brands once a video ad comes into view, recently Twitter and Yahoo have followed YouTube in waiting for at least a few seconds. Twitter and Yahoo each charge once a video ad has played for three seconds, and for its skippable TrueView ads, YouTube charges once someone has watched the full ad or at least 30 seconds of it.
Facebook's move may address one problem marketers may have with its autoplay video ads, but there's a bigger problem that spans the entire digital video ad ecosystem. All of these digital media companies offering different terms to advertisers creates a headache for marketers looking to suss out how much of a bang they got for their bucks. Marketers are already wrapping their heads around how the three-second-plus views they get on Yahoo or Twitter stack up in value to the full-ad views they get on YouTube, and now they would have to account for Facebook's 10-second-plus video ads and figure out how to standardize all these different measurements to determine how a campaign performed.
That struggle for marketers to assess their video ad buys on an apples-to-apples basis could hamper how effective Facebook's cost-per-view option may be.
"I don't know how consequential this is going to be purely because marketers say they want choice, but I really think they want consistency. But you need to test to find the right answer," Ms. Hofstetter said.
However, Facebook doesn't think its test of charging for 10-second views is the right answer. The company believes that if brands want to make sure people are watching the ads they buy, they should use Facebook's ad-buying tools. Those tools allow brands to tweak how many people they want to see an ad and how often, as well as aim the ads at people Facebook thinks are most likely to watch it.
"If I were advising my clients, which of course I do, my advice to them is to continue to use the reach and frequency tool and to continue to innovate and experiment with different video lengths," Ms. Everson said.
A danger of advertisers buying video ads based on 10-second views is that brands may place lower bids for these ads and therefore not reach as many people, according to Ms. Everson. "We believe that we are better off using our reach and frequency options, but we know that we have to be responsive to what clients are asking for and not just have one solution," she said.
Asked why Facebook chose the 10-second view threshold, Ms. Everson cited a Facebook-commissioned Nielsen study that looked at 173 campaigns that included Facebook video ads. The measurement firm found that increases in ad recall, brand awareness and purchase intent stem from the first few seconds people spend watching an ad.