Thibaut Aubin
Jun 17 · 7 min read

Nearly a year ago, we discussed on our blog the fundamentals of the brand safety concept regarding online advertising and the media industry. It brought up the point that content could be deemed “offensive” to some advertisers while acceptable to others, depending on the context and the industry each work in.

Yet with everything that has occurred so far in 2020, how do brands adapt their online advertising strategies to still be profitable and keep their brand image positive from web user’s perspectives when a large part of the news being published are rather negative? As we have seen, viewers increasingly consider the context of an ad’s appearance. With that in mind, as a brand, how do you select what you want to be affiliated with and why?


To get the real impact of this, it is first important to remind that most advertisers have little idea where their online advertisements are going to be placed on, and when. Their control over that aspect comes down to the information that is brought up by third parties, from Demand-Side Platforms on the selling side to Supply-Side Platforms on the buying side. The whole transaction happens almost instantaneously and most importantly programmatically, based on algorithms that are meant to associate advertisers to specific types of viewers on a publisher’s ad-space available based on criteria defined by the advertiser.

Image for post
Image for post
Simplified chart showing the different actors of online advertising

Whether you are looking to place an ad on a specific content or on the other hand trying to stay as far away from it as possible, it is crucial to have some knowledge over where your ads are shown so you’re in control of the messages you send out as an advertiser. In the past, we have seen instances where brands have lost tremendous amounts of money, several millions for the biggest of them, due to misplaced ads. Before they had secured algorithms, Google found themselves losing millions because major brands, such as AT&T, Walmart, Pepsi, Verizon and others pulled out their investments as their ads occasionally appeared on videos promoting hate speeches and extremist views.

Recent events that have transpired, from everything dealing with the coronavirus pandemic, the ensuing loss of jobs it has caused and economic recession, to now the protests following the agonizing death of George Floyd and the re-ignition of the Black Lived Matter movement, have caused brands and their advertisers to be very wary of where they place their ads and the content they want to be associated with.


Of course, it isn’t because most of these news have a rather negative aspect to them that no brand would want to be affiliated with them. The most obvious example would be brands and associations promoting medical help such as the American Health Care Association or the reduction of inequalities as a result of racism such as Color of Change or the Black Lives Matter Foundation would actually want to capitalize on these news to increase their performances and visibility, as they are perfectly tied to the news reports and the viewer’s interest at this given time. Other brands take advantage of these times to show solidarity to different causes in an attempt to transform their brand image as something more than just a corporation here to make money but also one that is compassionate and cares about building a better world. Anyhow, all brands would see an interest in making the best out of tough emotional content if they had control over their way to address the audience depending on the videos. When talking about the current social crisis in the U.S, there is a big difference between a video of a burning mall and a video of people marching peacefully in the street to protest for their civil rights.

Image for post
Image for post

Despite being viewed thousands or even millions of times, some of these “positive” videos did not receive the equal amount of advertisers they should have. Although it could attract great visibility, advertisers are not willing to take the risk yet to be associated to images of protests and looting or controversial remarks made by president Trump concerning the injection of disinfectant to combat the coronavirus for example. The possible repercussions of such affiliations could tremendously outweigh the possible gain in visibility and profit earned through the ad placed.


Innovations in artificial intelligence dedicated to the online video advertising industry are a true game changer for the way programmatic advertising will occur. Through the use of machine, but more importantly deep learning, if publishers and advertisers are able to understand exactly what the context of a video is and the elements taking place, deciding whether to display an ad or not, and where, will be much easier. Understanding the context of a video is being able to identify the different elements that characterise it. These can be the people that appear in the video, the setting it takes place in, what the emotions of the people seen in the video are, if any logos appear that could link a brand to the video, any potential violence or sexual content that needs to be taken into account etc…

Understanding the context of a video goes even deeper than being able to recognize the elements which characterize it. The real challenge is being able to make sense out of the elements recognized and most importantly having the ability to link different sets of data between them to obtain a holistic understanding of a video. For example, it’s one thing to recognize Rafael Nadal on a tennis court and the Nike logo on his clothing. It’s another to understand that Rafael Nadal is a professional tennis player, sponsored by Nike and therefore linked to that brand’s image through his actions, behaviour and performances on and off the court. Going even further, by understanding emotions in this example, brands such as Nike could adapt the ads they show according to Nadal’s emotions which testify of his performances.

As of right now, advertisers rely on the use of cookies to get a contextual understanding of the audience they wish to target and where to place their ads. Despite being great to build a profile of a given user to understand his preferences, what he likes to buy, his domains of interest and much more information, these lack efficiency to capitalize on a user’s instantaneous interest. Ads that are shown to the audience are still most of the time related to previous navigation, rather than current action of the user. In other terms, it is based on targeting the past more than understanding the present.


Having access to a technology enabling advertisers to understand in real time the context of a video about riots in Minneapolis or about the racial injustices happening in the United States and the consequences they have on the African-American population for example would change that paradigm. It would help them to programmatically capitalize on a specific context to convey a specific message, without the need for human moderators to check that the content is “brand safe” and that there won’t be a backlash to such action.

Thanks to this technological approach, issues related to brand safety will diminish through the use of programmatic advertising at scale. Advertisers will be able to set guidelines defining the content they want their ads to be shown on and will inquire much less risk of facing scandals such as the one Google had to face back in 2017. They will also have an easier time tracking where their ads were shown and why.

Image for post
Image for post
Some of the main events from 2020 that brands would be wary to associate themselves with. From top left to bottom right: Death of Kobe and Gianna Bryant; Covid-19 virus and the crisis it has caused; Australian wild fires; death of George Floyd; Biggest crash of the Dow Jones since 1987; UK officially withdraws from the EU; Donald Trump impeachment trial

Reminiz has been tackling this problematic for years now and possesses a technology able to do just that. Trusted by several top-10 media groups in the United States and France, its artificial intelligence can recognize celebrities with unmatched results on the market as well as surroundings a video may take place in, logos that may appear or even actions occuring, such as presence of explicit content. But again, brand safety here can only prove its value if technology is able to make sense out of content. The nature of explicit content is as important as the context of appearance. Advertisers might judge acceptable the presence of underwears in a comedy, but much less in any User Generated Content.

Great progress has been made in recent years concerning brand safety and a lot less mistakes occur, yet it still relies too much on human moderation or just random elements to select who an ad will be shown to, why, and when. Soon enough brands and advertisers will have increased confidence in their ad placements thanks to artificial intelligence having been developed. Judging by the nature of dramatic events in 2020, one could judge that timing could not be better.

Context Insights

Knowledge Builds Trust

Medium is an open platform where 170 million readers come to find insightful and dynamic thinking. Here, expert and undiscovered voices alike dive into the heart of any topic and bring new ideas to the surface. Learn more

Follow the writers, publications, and topics that matter to you, and you’ll see them on your homepage and in your inbox. Explore

If you have a story to tell, knowledge to share, or a perspective to offer — welcome home. It’s easy and free to post your thinking on any topic. Write on Medium

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store