Brand safety: The missing manual for publishers
Like many other ad tech buzzwords—the definition of brand safety is constantly evolving and varies depending on whom you ask. Since brand safety is mainly considered a buy-side issue, it is frequently covered from the perspective of marketers, brands, and advertisers.
However, publishers are directly affected by brand safety, too, sometimes as a broader industry participant and sometimes individually, as advertisers reduce their advertising spend and use keyword blocklists to better suit their campaign goals and brand safety needs.
In this post, we’ll cover brand safety from the publisher lens, including what brand safety means, how it is enforced, the cost of brand safety for the sell-side, and what publishers can do about it.
What is brand safety?
Brand safety refers to a set of measures that brands implement to prevent their advertising from finding placements in unsafe or inappropriate editorial environments. Most brands tend to avoid the so called “dirty dozen” categories, which includes military conflict, obscenity, drugs, tobacco, adult, arms, crime, death/injury, online piracy, hate speech, terrorism, and spam/harmful sites. The Interactive Advertising Bureau (IAB) also added a 13th category to the list: Fake news.
While it makes sense that marketers are driven to protect their brand image and reputation by staying away from objectionable content, the current breed of brand safety tools are part of an emerging technology and therefore tend to create many false positives. This in turn, disrupts the existing ad-supported model of the Web, especially for large news publishers reporting on hard news, by making it harder for them to effectively monetize all their webpages.
How brand safety is enforced
Brand safety vendors will scan the publishers’ inventory and webpages for multiple signals and assign a “safe” or “unsafe” label to them. Based on this label, the advertiser will either place a bid (if the inventory is tagged safe) or not participate in the auction (if it is tagged unsafe).
Here are a few factors that brand safety vendors consider when evaluating a publisher’s inventory and how advertisers enforce their brand safety protocols:
Invalid traffic
IVT is a broad term that encompasses any activity that doesn’t come from a real user with genuine interest. This includes accidental clicks, fraudulent clicks, incentivized clicks, advertising botnets, and more. IVT is also correlated with malicious scripts and data breaches, making it a security threat for users. The presence of IVT signals to buyers that the publisher isn’t doing their part in keeping their inventory clean for advertisers and safe for users, which is why most brand safety vendors will stop their ads from appearing in high-IVT environments.
Keyword blacklists
Keyword blacklists contain stop words linked to the “dirty dozen” content categories mentioned previously. These blacklists are maintained by DSPs and brand safety vendors and if a word matching the blacklist is found on the publisher’s webpage, the ad serving is terminated.
Blacklists are a blunt tool and don’t take context into account. They are also ballooning, with some lists containing as many as 3,000 to 4,000 words. “An unintended consequence is that perversely some of the best and most trusted content is being blocked inadvertently,” said James Wildman, chief executive of Hearst UK, in a Guardian article. “If we write about the Duke and Duchess of Sussex [Harry and Meghan], chances are most software would see the word ‘sex’ and assume it is inappropriate content to appear alongside and block it. We are getting blocked for keywords like ‘shoot’. We write about photoshoots constantly.”
Site whitelists
Whitelists are an inversion of the idea behind keyword blacklists. Some buyers want to exercise tighter control than what keyword blacklists will allow, so they go one step further and specify the exact list of websites where their ads can run.
“Luxury advertisers, for example, often operate whitelists featuring between five and eight publishers or between 10 and 15 sites,” said Emily Brewer, head of publishing at Teads in a Digiday article. “With such limited scale, and as advertisers layer on more targeting criteria and attempt to avoid appearing next to their competitors, this audience quickly dwindles.”
Compounding the problem for publishers is that fact is that they have no way of knowing if they have been placed on a whitelist or blacklist. Moreover, since these lists are built manually and not usually updated dynamically using any logic or algorithm, once a publisher is on the list, they might stay on or out of it for a long time.
Fake news
Fake news counts as sites that are creating and distributing news or information that is factually inaccurate. Entities that engage in publishing fake news rely on shocking revelations or incendiary language to rack up traffic and pageviews.
In cases where a controversial headline is used, brand safety vendors can easily assess and mark the content “non-safe” by conducting a keyword analysis. However, the other category of fake news, where seemingly legitimate content is not true or misleading, is harder to identify.
The rise of fake news has promoted the creation of independent fact-checking entities such as Snopes, which helps users fact check news and also partners with platforms for real-time verification of trending content. If buyers or ad tech vendors acting on behalf of them notice a publisher routinely publishing fake news, that becomes a easy reason to end up on a blacklist.
The cost of brand safety for publishers
There are both individual instances and wider estimates of how brand safety concerns and the technology used to enforce it has led to reduced ad spend and less revenue for publishers.
In 2017, after noticing its ads being served next to offensive content, JPMorgan Chase decided to cut the number of sites that it advertised on from 400,000 to just 50,000. Around the same time, consumer goods brand P&G slashed its digital ad spending by $140 million over similar concerns. There are valid reasons for why brands take such a proactive approach to safeguarding their reputation. A CMO Council study confirmed that almost half of consumers will rethink purchasing from a brand after seeing their ads next to objectionable content.
According to a Guardian report, UK publishers lost £170m in ad revenue in 2019 as brand safety tools that stop ads from appearing on hard news topics, such as shootings and terrorism, also inadvertently blocked them on other popular stories that do not fall under that category.
More recently, even ads on the homepage of reputed publications such as New York Times and WSJ were blocked last year, only because the page mentioned “covid-19” or “coronavirus” somewhere. “The result of this blocking is that more ads and dollars flow to lower quality sites in programmatic channels – the exact opposite effect to what advertisers thought would happen,” independent ad fraud researcher Augustine Fou said in his Forbes column.
So how imprecise are brand safety tools? Adalytics tried to find out with their research. “This exploratory study shows how an estimated 21% of economist.com articles, 30.3% of nytimes.com, 43% of wsj.com, and 52.8% of articles on vice.com are being labeled as “brand unsafe”,” the study revealed. “It also illustrates how journalists who focus on certain “serious” topics, such as Middle East affairs, obituaries, or political events, are disproportionately likely to have their work marked as “unsafe” by brand safety vendors.”
While there isn’t a lot publishers can do in response to how brand safety tools categorize their content, and they definitely can’t stop reporting hard news, it’s important to keep a clean house and minimize the chances of being penalized due to reasons that can be easily fixed.
So, what can publishers do?
Keep your inventory clean
Most advertisers and the ad tech vendors who act on their behalf monitor publisher inventory both before and while running campaigns for signs of IVT and suspicious activity. It is therefore imperative for publishers to keep a close eye on where their traffic is originating from, in order to catch and resolve anomalies before they turn into problems. Refer to this article on monitoring and preventing IVT for more information. Publishers should also consider partnering with an analytics vendor who can assist them in ensuring that the inventory is clean and safe for users.
Only work with trusted partners
Publishers need to work with a number of technology vendors to enable content management and measurement, site analytics, login systems, and monetization. Many of these vendors will require publishers to insert their third-party integration code in order to work, and specifically in case of ad tech vendors, adding a line in the ads.txt file may be required. These integrations can potentially create the technical conditions that allow for malicious activity such as security breaches, domain spoofing, and inventory arbitrage. This is why it’s crucial to work only with trusted vendors, avoid adding too many “RESELLER” entries into the ads.txt file, and periodically monitor third-party code and ads.txt file to remove redundant and unused entries.
Improve your viewability score
Viewability in ad tech is a measure of how many ads are actually being seen by human users. According to the Media Ratings Council (MRC), a display ad is considered viewable if “at least 50 percent of the ad’s pixels are in view for at least one second”, based on how many of the total ads served on page meet that threshold, publishers can calculate their viewability score. A viewability score above 60% is considered good, but premium publishers tend to have scores well above that. There can be multiple reasons for a especially low viewability score, including IVT, suboptimal ad placements, and technical ad serving issues. Advertisers, DSPs, and brand safety vendors will either avoid bidding on or reduce bids on inventory with low viewability score. Publishers must therefore monitor and undertake initiatives to improve their score.
Partner with a brand safety vendor
The brand safety vendors that work with advertisers and DSPs also work with publishers and advise them on ensuring that their inventory is meeting the industry brand quality standards. By working directly with brand safety vendors, publishers can also make more informed decisions about their monetization options. For instance, a hard news publisher might be able to identify that a certain content category is consistently blocked by brand safety vendors, allowing them to pivot to making direct deals for or running contextual advertising on affected webpages.
Offset losses with alternate revenue
Despite all publisher efforts, it may not be completely possible to mitigate all the revenue loss incurred by the current and future versions of brand safety tools used by advertisers. The smart play is to find and activate alternate revenue streams to offset the potential losses. There are a lot of techniques that can help publishers grow their incremental revenue, including header bidding, ad refresh, experimenting with subscription, and adblock revenue recovery. Depending on the existing adblock rate of the website, ad recovery via Acceptable Ads is a easy and efficient way to unlock the advertising revenue that publishers may be losing due to adblocking software. Use our adblock revenue recovery calculator to generate an estimate for your website.