About three years ago, YouTube found itself at the centre of a brand safety crisis when The Times found ads shown against jihadi propaganda films. Cue a slew of brands pulling spend from the platform, much finger-pointing, and a very public apology from YouTube.
It’s every marketer’s nightmare… and while Nestle may have avoided being caught up in Jihadigate in 2016 it found itself in a comments crisis last week when its ads were shown against kids videos, which carried a slew of comments linking to paedophilic content.
So how has this happened again, what should be done, and by whom?
YouTube will inevitably cop the most flack. As the world’s most-viewed and deepest-pocketed video platform, by a mile, and it should lead in terms of standards, rather than trudge grudgingly behind.
YouTube, like Facebook, Snap, Twitter and all the tech giants, sit in a regulatory twilight zone between platform and publisher, and until the government formulates its wider policy on big tech - this hybrid video-cum-social network is going to remain prone to those who want to use the service for unlawful means. Even Facebook accepts the case that it should be regulated. But until that happens, it’s down to brands, agencies, and media owners to find a way, and it’s not easy.
For our part, as an agency, brand safety is already a complicated issue. Different clients have specific requirements, and one brand’s no go may be a keyword for another. And there’s a constant need to review the blacklists, keyword and phrase exclusions which we have put in place.
Clients have to make a difficult trade-off between quality and quantity. That’s how all markets work. But guaranteed brand-safe environments come at a premium, which means for us delivering performance targets on a plan for the right price, clients either need to pay a higher premium or accept reduced accuracy. As a performance specialist I think we’re good at finding the balance and it’s in our culture to encourage clients to pay a premium in pursuit of quality, but it’s not always an easy conversation to have; and one that I suspect may never be had if performance and verification isn’t particularly high on a client’s list of priorities.
Finally, what about the responsibility of the media owner? To put things in perspective, with an estimated 300 hours of video content being uploaded every minute, pre-moderation of video content (let alone comments) simply isn’t plausible. Nor is it possible to guarantee that every video can be 100% brand-safe for every client, that tags are correctly applied by the uploader (how many of those jihadi videos were tagged as ‘automotive’ or ‘travel’ to attract the highest number of bids) - even if the agency (or in-house team) is using the tools correctly.
So there’s a debate to be had about how extensively YouTube should reactively moderate (witness Facebook’s increasing investment in human moderators) or invest more in algorithms that can identify suspicious comment threads.
A reduction in addressable inventory would mean YouTube’s revenue decreases, they take a hit on margin, or - more likely - the cost is passed on to advertisers. But compared to the fines, the bad PR that affects us all, perhaps that’s a price worth paying, given the fact that Disney, Nestle, McDonald’s and AT&T all pulled ads from the platform this week despite a lack of clear evidence of the scale of the problem; with an estimated 5bn digital impressions traded in the UK per quarter by the major networks alone… there are bound to be brand safety mistakes - but it’s a perception that matters and big brands are risk-averse.
It’s clear that the brand safety conversation shouldn't stop at the addressable video and as an industry, we need to pay more attention to contextual comments.
Brand safety is a topic we’ve always prioritised at agenda21; we were the first media agency to become both DTSG Brand Safety and Ad Fraud certified by the industry body JICWEBS. Our co-founder Pete Robins has even contributed to the brand safety guidelines that now the majority of the industry follows.
The problem, however, is that JICWEBS remit extends only to the guidance around biddable media and the brand safety controls in the immediate context - but no further. In other words, pre-roll and mid-roll video yes, comments and the links in comment sections, no. And, for all the good work that takes place within the IPA and JICWEBS, YouTube is but one of many contributors and brand safety is but one item on a long agenda including ad fraud, viewability, transparency and many many more.
Which brings to where I started. In a rapidly changing digital market, I think we all have to accept that we will make mistakes, resist the urge to point fingers and work more transparently and collaboratively. But I think that starts with even closer collaboration with YouTube specifically, given its scale and market impact. After a job well done on video content, maybe the focus now needs to be diverted into new areas for everyone’s benefit.