By now I am sure most people have come across any number of articles about advertisers pulling spend from Google & YouTube due to brand safety concerns. Most of them gloss over the details on exactly how these incidents have come about and very few of them clearly and simply explain what the problem is.
What is the big deal?
The crux of the issue is that a number of advertisers particularly in the UK have been found running their ads before hateful and extremist videos on YouTube. Clearly this is far from a good look for anyone. Subsequently Google has been unsuccessfully trying to say they have the issue under control and talks between Google and the advertisers have basically broken down. The advertisers response has been to pull spend and force Google to act.
How did that even happen?
To understand that we need to look at how content becomes eligible for advertising with Google.
- The content owner needs to sign up for Google Ad Sense (Google’s program to sell ads)
- They setup an account and then implement tags or allow ads on their YouTube channel
- Google uses algorithms to categorise the content and opens it to ads
- Any ‘risky’ content is flagged and reviewed manually
Steps 3 and 4 is where things get a bit messy for Google, simply put they need to process and understand hours and hours of video uploaded to YouTube every second of every day. Naturally things can fall past their filters especially as it is particularly hard for an algorithm to understand the content of the video (consider this is essentially writing a math formula to watch and understand a video).
Advertisers end up on bad placements when they first join ad sense, as there is a period between the review process that content is allowed to sell ads but has not been manually reviewed.
Sounds scary, what should we do?
Fortunately if you are currently working with Starcom and its AOD team then you don’t need to be concerned, before this all came about we already had strict brand safety process in place. The primary method we have used is to run only on a selection of websites reviewed by our team. This means any AOD video and display activity can’t serve on any new sites unless they have been reviewed by our team.
We however are working closely with Google as a global network to push them to develop better systems to ensure YouTube is brand safe and will continue to implement best in class process to ensure all clients are protected.