The perils of online advertisements

When Google acquired YouTube in 2006 for $1.65 billion, it was considered a pricey gamble, one made with the belief that an online service known for pirated videos and vapid user-generated content could appeal to major advertisers.

The bet paid off. YouTube is now one of the pillars of Google’s advertising business and the most valuable video platform on the Internet. In recent years, advertisers, unable to ignore its massive audience, flocked to YouTube to reach younger people who have started to shun traditional broadcast television. But the technology underpinning YouTube’s advertising business has come under intense scrutiny in recent days, with AT&T, Johnson & Johnson and other deep-pocketed marketers announcing that they would pull their ads from the service.

Their reason: The automated system in which ads are bought and placed online has too often resulted in brands appearing next to offensive material on YouTube such as hate speech. Recently, the ride-sharing service Lyft became the latest example, removing their ads after they appeared next to videos from a racist skinhead group.

“This is beyond offensive,” a Lyft spokesman, Scott Coriell, said. “As soon as we learned of it, we pulled our advertising on YouTube.” The pullback from advertisers strikes at the core of YouTube’s appeal. Unlike television, with specific programming during which brands choose to run their advertising, YouTube mirrors the Internet’s sprawl, specialising in niche content that may not appeal to a mainstream audience but attracts engaged viewers. This provides YouTube with an enormous audience watching one billion hours of videos a day, perfect for new ad technology that minutely slices and dices an audience so that companies can target specific viewers.

Use of automation

That technology, known as programmatic advertising, allows advertisers to lay out the general parameters of what kind of person they want to reach — say, a young man under 25 — and trust that their ad will find that person, no matter where he might be on the Internet. This approach plays to the strengths of tech giants like Google and Facebook, allowing advertisers to use automation and data to cheaply and efficiently reach their own audiences, funneling money through a complicated system of agencies and third-party networks.

But more than 400 hours of content are uploaded to YouTube every minute, and while Google has noted that it prevents ads from running near inappropriate material “in the vast majority of cases,” it has proved unable to totally police that amount of content in real time.

“The simple truth is that the same tech that allows the posting of a recipe, the joyous video of a child or the exposure of an excessive act from law enforcement allows the creation, posting or sharing of the video of a murder,” Rob Norman, chief digital officer of GroupM, the media buying arm of WPP, wrote this week in Campaign, a trade publication.

Marketers have seen programmatic advertising as a groundbreaking development in media — a technologically efficient way to leverage the expanse of the Internet so that, for example, Pampers can reach a new mom on a local blog or an instructional video about how to deal with a newborn’s baby acne.

This has opened up the whole of the Internet to brands, which typically opt to remove their ads after they have appeared on unsavory sites, rather than restrict the ads to a list of preapproved locations.

That sheer scale, coupled with its reliance on algorithms rather than humans to filter out the objectionable content after it appears, has been Google’s main defense. More than two million websites are a part of its display advertising network.

“What we do is, we match ads and the content, but because we source the ads from everywhere, every once in a while somebody gets underneath the algorithm and they put in something that doesn’t match. We’ve had to tighten our policies and actually increase our manual review time, and so I think we’re going to be OK,” Eric Schmidt, chairman of Google’s parent company, Alphabet, said.

While brands have expressed concern about showing up next to unsavory photos and videos uploaded to digital platforms by users, the situation with YouTube is particularly jarring. YouTube splits advertising revenue with its users, meaning advertisers risk directly funding creators of hateful, misogynistic or terrorism-related content.

There’s a lot at stake for YouTube and Alphabet. Search advertising is not growing as fast as it once did, and television ad budgets are still larger than total spending on digital advertising. YouTube and Facebook have made no secret of their desire for television advertising funds. James Dix, a senior media analyst at Wedbush Securities, estimates that YouTube accounted for revenue of about $13 billion, or roughly 20% of all advertising revenue on Google’s Internet properties in 2016.

“If this company is going to look for new growth at scale, it has to pull money from television,” Dix said.