Data, AI and Brand Safety: Challenges and Opportunities

GUEST OPINION: The proliferation of fake news, toxic digital environments and the demise of the cookie means for marketers, brand safety and suitability has become a key focus for engaging with customers and building trust.

To help monitor where marketing and advertising content is shown, brands are implementing AI-enabled tools to place their ads in ‘brand-safe’ contexts. But it’s clear that a ‘one size fits all’ solution isn’t enough for today’s climate.

Unfortunately, the vast majority of traditional contextual solutions were built at the dawn of programmatic advertising, and many have not kept up to speed with the nuances of how harmful online content is framed, worded, and produced.

The data algorithm and keyword targeting

Dated algorithms rely on keyword targeting and blocking, which alone do not protect a brand entirely. And, with restrictions around the walled gardens which makes it even more difficult to manage brand safety in today’s polarised environment, businesses are rightfully searching for new solutions.

In 2020, the unforeseen global pandemic brought brand safety and suitability challenges into sharp focus for many marketers.

As publishers across the globe unveiled the impact of COVID-19 to their readers, they simultaneously began to lose revenue from a broad spectrum of advertisers who were blocking pages inadvertently.

The fundamental methodology of many brand safety tools relies simply on blocking keywords. For example, if an advertiser decided to add COVID-19 to their blocklist in order to avoid appearing next to negative news stories, they would effectively be blocking almost all of the leading global news sites of the time.

A recent Ofcom report showed nearly nine out of 10 adult internet users turn to traditional media as a source of COVID-19 information. Hence advertisers could be missing out on an abundance of audience engagement opportunities.

This is backed by a study commissioned by CHEQ and Digiday, which reports nearly two-thirds of advertisers stating brand safety tools are not fit for purpose. Approximately 92% of marketers stated they would forgo the use of brand safety tools if they were not achieving adequate reach, while 99% are seeking more customised tools to ensure safety, without sacrificing reach. Brands are at significant risk of both long and short term consequences as a result of forgoing brand safety for reach.

AI and the new era of contextual targeting

Marketers need to expand their toolkit and rethink the concept of ‘brand safety etiquette,’ and embrace suitability targeting in order to avoid the risk of missing the mark on the context in which their message is placed. This is where the latest AI-enabled contextual outcome engines can help marketers have greater visibility and understanding of nuances and true context, and allow them to identify the right moment and the right environment to align with their message.

New contextual targeting innovations can now help brands easily create brand-safe environments, ensuring ads appear on the correct sites, which are not negatively associated with toxic messages or harmful environments. As an example, brands can now use tools to build contextual understanding when looking for specific words, such as identifying words that may seem dangerous but are not harmful in the context of where they are placed.

Brands can now adopt a variety of contextual tools to safeguard them against future threats. These tools must define the perfect context for their ads to appear against, and to apply that context across all addressable channels, including display, video and mobile, programmatic audio, addressable TV, and in-app.

To address contextual video, for example, marketers need a computer vision-based solution that can identify objects, logos and nudity among various signals. This allows the platform to seamlessly offer sophisticated targeting and brand protection to this premium inventory.

An open architecture approach?

Modern marketers who want to thrive in the new marketing age need to be aware of nuance and the true context of the content to identify the right moment and the right environment with which to align with the relevant message.

Brands looking for safe and suitable martech solutions should also align themselves with advanced and trusted products that move far beyond the scope of traditional brand safety and targeting methods.

Lastly, organisations should have an independent, open-architecture approach in their chosen contextual outcomes engine to ingest first-party data from the implemented platform. From DMPs and CDPs to DSPs and ad servers, to be able to create bespoke contexts and apply them to future campaigns.

Real example – 4D implementing DeepSee

In April 2021, context outcomes engine 4D saw the addition of DeepSee to its Dimensions Marketplace to enhance targeting and brand suitability context for its clients beyond traditional block-listing. DeepSee works by identifying and protecting precious brand dollars against today’s site fraud by providing quality scores for domains.

Using AI to examine the nature of websites and their networks to discover places that harbour site fraud, DeepSee’s analytics doesn’t rely on outdated predictions by solely collecting large amounts of data from web users. This helps marketers determine whether they should trust a site to run advertising.

By expanding its ‘Dimension Marketplace’ with the addition of DeepSee, 4D enables advertisers to further ensure their ads are running on suitable content and their customers have positive interactions with their advertising.

Click here to view original web page at

By admin

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.