Last week, UK advertisers, including the government, the Guardian newspaper, and various others began boycotting Google’s ad products including YouTube over the fact their ads were appearing next to troublesome content, ranging from videos promoting hate to those advocating terrorism. Unsurprisingly, given the exact same issues exist here in the US, the boycott this week began to spread to Google’s home turf, with several of the largest US advertisers pulling their ads from some or all of Google’s platforms. The challenge facing Google is this problem has no easy fix – with two of three possible scenarios, either creators or advertisers will be unhappy, while Google is probably hoping a third scenario is the one that actually pans out.
The main focus of the complaints has been YouTube, although the same problem has, to some extent, affected Google’s ads on third party sites as well. On YouTube, the root of the problem is the site has 400 hours of video uploaded every minute, making it impossible for anything but an army of human beings to view all the new content being put onto the site continuously.
As such, Google uses a combination of algorithmic detection, user flagging, and human quality checking to find videos advertisers wouldn’t want their ads to appear in and those systems are far from perfect. Terrorist videos, videos promoting anti-Semitism and other forms of hate, content advocating self-harm and eating disorders, and more have slipped through the cracks and ended up with what some perceive as an endorsement from major brands. Those brands of course, aren’t happy with that. Following some investigations by the UK press, several have now pulled their ads either from YouTube specifically or from Google platforms in general until Google fixes the problem. US brands like AT&T, Verizon, Enterprise, GSK, and others are starting to follow suit.
No Perfect Fix
From Google’s perspective, the big challenge is its existing systems aren’t working and there’s no easy way to fix that. Only one reasonable solution suggests itself and it’s far from ideal: restrict ads to only those videos which appear on channels with long histories of good behavior and lots of subscribers. That would likely weed out any unidentified terrorists, hate mongers, and scam artists without having to explicitly identify them. Problem solved! Except that, of course, the very long tail of YouTube content and creators would be effectively blacklisted even as this much smaller list of content and creators are whitelisted. That, in turn, would be unpalatable to those creators, even if advertisers might be pacified. Of course, it would have a significant effect on YouTube’s revenue too.
Given that some creators are already unhappy with what they see as the arbitrary way YouTube already determines which videos are and aren’t appropriate for advertising, going further down that route seems dangerous and will create problems of its own. But, given the current backlash against YouTube and Google more broadly over this issue, it can’t exactly keep things as they are either, because many advertisers will continue to boycott the platform and there’s likely to be a snowball effect as no brand wants to be seen as the one that’s OK with its ads appearing next to hate speech, even if others aren’t.
So, we have two scenarios, neither of them palatable. One would be essentially unacceptable to the long tail of creators and would likely significantly impact YouTube’s revenue, while the other would continue to be unacceptable to major advertisers and also would significantly impact YouTube’s revenue. To return to a point I made at the beginning, this actually is broader than YouTube to programmatic advertising in general, including Google’s ads on third party sites. Alphabet’s management has cited programmatic advertising, where humans are taken out of the picture and computers make the decisions subject to policies set by site owners and advertisers, as a major revenue driver in at least its last four earnings calls, mentioning it in that context at least seventeen times during that period.
To the extent the programmatic method of buying is a major source of the content problem at YouTube specifically and Google broadly, that’s particularly problematic for its financial picture going forward. There was already something of a backlash over programmatic advertising towards the end of last year when brand advertising was appearing on sites associated with racism and fake news but this YouTube issue has taken to the next level.
Hope of a Third Scenario
Alongside these two unappealing scenarios, there’s a third. Google must be hoping this one is what actually pans out. This third scenario would see Google making more subtle changes to both its ad and content policies than the ones I suggested above and eventually getting advertisers back on board. That approach banks on the fact brands actually generally like advertising on Google, which has massive reach and – through YouTube – a unique venue for video advertising that reaches generations increasingly disengaged from traditional TV. So I’d argue advertisers don’t actually want to shun Google entirely for any length of time and mostly want to use the current fuss to extract concessions from the company both on this specific issue and on the broader issue of data on their ads and where they show up.
Google’s initial response to the problem, both in a quick blog post on its European site last week and a slightly longer and more detailed post on its global site this week, has been along these lines. It’s accepted responsibility for some of its past mistakes, identified some specific ways in which it plans to make changes, and announced some first steps to fixing problems. However, the fact that several big US brands pulled their advertising after these steps were announced suggests Google hasn’t yet done enough. It’s still possible advertisers will come around once they see Google roll out all of its proposed fixes (some of which were only vaguely described this week) and perhaps after some additional concessions. That would be the best case scenario here. Some of the statements from advertisers this week indicate they’re considering their options and reviewing their own policies, suggesting they may be open to reconsidering.
But these current problems still highlight broader issues with programmatic advertising in general, on which advertisers won’t be placated so easily. I could easily see the present backlash turn into a broader one against programmatics in general, which could slow its growth considerably, with impacts both on Google and the broader advertising and ad tech industries. I would think Google/Alphabet would be extremely lucky to emerge from all this with minimal financial impact and I think it’s far more likely it sees both a short-term dent in its revenues and profits from the spreading boycotts and possibly a longer-term impact as brands reconsider their commitments to programmatic advertising in general.