Keyword block lists still cause headaches for publishers
While it’s been a few months since the last wide-scale media furor around a brand’s ads showing up next to unsavory or objectionable content, publishers are being penalized more than ever over blunt brand-safety tools that rely on lengthy keyword blacklists.
The issue is that many clients still apply generic keyword blacklist strategies across all the publishers they work with. Keyword lists and brand-safety tools can be a woefully blunt and ambiguous: A single word, without the right context, can misrepresent the meaning of the article, leading clients to restrict ads from running on it. Sports writers invariably use the word “shoot” for scoring a goal, but publishers find this word is often on these lists due to connotations with violence, demonetizing any content featuring it.
Sometimes, the outcomes of these keyword block lists border on the absurd. Paul Wallace, vp of media solutions at Vice Media, recounted at this week’s Digiday Programmatic Summit how a health article about constipation was dinged for brand safety because the word “poop” was in the headline: “How Many Times A Week Should You Be Pooping?” Defecate would be OK, but poop was a problem. In another instance, a Vice article about the window for use of Plan B was flagged because “screwed” was used in the headline, “You’re Screwed If You Wait Too Long to Take Plan B.”
The examples are numerous. Hearst, which had 24.3 million unique monthly users in October (according to Comscore) across titles like Elle, Esquire and Cosmopolitan, began vetting the keyword lists and articles blocked from its brand-safety vendor, said Ryan Buckley, head of programmatic. It found that all articles featuring Meghan Markle, the Duchess of Sussex, had been blocked across its portfolio for containing the word “sex” despite the articles not violating brand-safety guidelines.
Another example from Hearst: An article with “makeup” in the URL and the article ID containing the number 47 was stitched together to arrive at the firearm, AK-47.
“The fact is no marketer or agency has got sacked for blocking too many words, but they certainly could if their ad is seen on dodgy content, so they just load on more,” said Bedir Aydemir, head of audience and data, commercial at News UK. “It basically means they are blocking ‘news.’ Where on earth do you think your ads are appearing if you are blocking thousands of words? I can only imagine it must be recipe sites.”
Unduly blocking articles featuring Meghan Markle that otherwise would be brand-safe is one thing. Another is the general nervousness that clients feel about advertising next to news, which, it’s assumed, will lead to knock-on negative effects for the featured brand in the ad next to an article about Donald Trump or Brexit. For this, tech providers are moving the narrative from brand safety to brand suitability, according to publisher sources, which will add more nuance to the environment brands want to appear in.
Studies by measurement firms and trade bodies, like Comscore and Newsworks, show that news content improves ad effectiveness, through measures like ad recall, and doesn’t have such a negative effect on the brand even if the content is hard or challenging. But advertisers, looking for quick efficiencies, often end up over-blocking. While this makes it harder for agencies to meet impression thresholds for campaigns, it’s not the existential and revenue threat that it is for publishers.
“A publisher told me that they felt that keyword blocking had a higher impact on revenue than GDPR,” said an ad tech exec.
Previously, publishers have reported unwarranted blocks 30% of inventory up to 90%, but unless there’s a direct relationship between the buy and sell side, it’s hard to gauge the real impact if clients are using pre-bid blocklists on the open marketplace.
To get a better understanding, Hearst is asking vendors to provide lists of what they are seeing blocked across the portfolio so it can scrutinize the blocked article URLs itself.
“The challenge is we don’t get that visibility as publishers,” said Buckley. “In a programmatic world, we don’t see what level of blocking is occurring across our sites. The only opportunity we have is when we have that direct relationship with the advertiser and agency to really identify what is being blocked. It’s the industry standard to use brand-safety companies, but we need to challenge those and make sure they are taking into account sentiment, context and tone.”
Brand-safety companies are ensuring publishers they are bringing more nuance to brand-safety tools. This week, ad verification company Integral Ad Science announced the purchase of contextual analysis company ADmantX, a move welcomed by publishers to further signal the move away from probabilistic artificial intelligence keyword blocking towards more natural language processing technology and context.
In October, Daily Mirror publisher Reach said 40% of traffic across news, sports, technology and celebrity-related articles was actually brand-safe, but it wasn’t monetized. To counter this, it developed a tool, Mantis, using IBM Watson’s machine learning, natural language processing to identify articles that are safe but have been blocked. Publishers like The Telegraph as well as agencies are talking with Reach about how they can use Mantis.
But publishers using these tools will only go so far if it’s not adopted by agencies and advertisers. Against that backdrop, it’s only a matter of time before another brand-safety blunder will force agencies to jack up the brand-safety dials.
“This topic will explode again; they [advertisers] will block everything,” said Buckley. “There’s a knock-on effect for agencies how to you reach the impressions if there’s are 3,000 keywords blocked.”
“This is an issue that clients need to fix, and I don’t think they have any incentive to do so,” added Aydemir.
The AMERICA Act spotlights Capitol Hill’s ingrained antipathy for Big Tech
A reprised version of the Competition and Transparency in Digital Advertising Act spells trouble for double-sided marketplaces.
How BuzzFeed’s Creator Score is grading the impact of its creator network
BuzzFeed's Creator Network is a primary focus in 2023 for the publisher, and its campaign grading tool is being used to prove out its ability to create successful ads.
In graphic detail: Google’s Ads Safety Report shows suspect ad activities are on the rise
Google's ad transparency efforts detail how bad actors necessitate further investment.
SponsoredHow critical data pillars will increase brands’ confidence in CTV
This article is part of Digiday’s coverage of its Digiday Publishing Summit. More from the series → Mario Diez, CEO, Peer39 With every quarter, the balance of TV viewership slips away from the traditional linear model and more towards connected TV. Less than half of the adults in the U.S. subscribe to cable or satellite, […]
Media Briefing: Publishers share their biggest challenges and opportunities at the Digiday Publishing Summit
While Q1 ad revenue, sales cycles and payment windows appeared to be equally bad across the media industry, bright spots arose around consumer revenue streams, new tech experimentation and traffic patterns.
How Dotdash Meredith worked through the challenge of integrating two major publishers’ tech stacks
At the Digiday Publishing Summit, Dotdash Meredith product chief Adam McClean shared why he had initially expected the integration to be done by summer 2022 and why the work is still ongoing.