Breaking Down Bad Ads: User Experience & Ad Quality
The State of Ad Quality
Bad ads sources
They are equally as likely to see bad ads coming through the header or S2S than from any other sources.
Low-quality ad creative
Unwanted ad creative is a more widespread problem in the header than redirects are, 52% say they’ve had to deal with low-quality creative in the header, 23% with redirects in the header.
Header security risks
Bad ads of any variety — security or quality risks — count as one of the greatest challenges in the header for 48% of respondents.
What is a Bad Ad?
Ad security is focused on malvertising and malware prevention, which is so concerning and immediate to publishers and platforms that it warrants its own discussion.
Bad Ads: Beyond Security Risks
Managing Ad Quality
Bad Ads & User Experience
How User Experience Effects Revenue?
Types of Bad Ads
1
Deceptive content or misinformation
2
Pornographic, sexually suggestive or otherwise “adult” content
3
Fraudulent or non-existent products
4
Gambling sites, cryptocurrency, multi-level marketing networks, or other financially risky ventures
5
Competitors’ ads
6
Disruptive, heavy, or otherwise off-spec ads
Among the category of deceptive ads,
we see three types of deception:
Type 1:
The landing page content is not related to the ad content at all.
Type 2:
The landing page is related to the ad content, but the page contains low-quality content (misleading, off-brand, etc.).
Type 3:
The ad leads to a phishing scam, a malware download, or other unsafe content.
Read our malvertising guide for more details on how malvertising compromises user security
Ad Quality During Covid-19
Since the beginning of the COVID pandemic, publishers have seen an explosion of bad ads — including deceptive ads, brand-unsafe ads, ads for phantom or substandard products, and predatory (i.e. surge-priced) products. According to a report Digiday and GeoEdge produced, 67% of publishers have encountered a significant or great amount of deceptive ads since the spring of 2020. Among those deceptive ads, their most prevalent concern is ads that lead to landing pages unrelated to the ad content itself. This is particularly concerning because so many deceptive ads today use clickbait methods — sensationalistic copy or images, or tabloid-style unauthorized celebrity images. Such ads don’t only make the page look cheap; they have unusually high CTR. The landing page must be a consideration for QA.
Where Bad Ads Come From?
Ad Cloaking
One common — and extremely difficult to detect and block — method of disguising is to use cloaking. Cloaking attacks hide the code for a malicious ad creative within code that looks legitimate or innocuous. Cloaking also hides the URL for a malicious landing page within code that resembles a URL for a legitimate site. In some cases, malvertisers will point the fake URLs to sites that counterfeit the design and branding of legitimate sites. Cloaking attacks detect whether the ad is in a non-human environment (an ad creative scanner) or a human environment, and will show the fake creative and URL to a scanner. When the page loads, the real (malicious) creative and URL are swapped in and shown to the human user. Cloaking can only be properly detected by sophisticated, real-time blocking technology.
How Bad Ads Enter The Supply Chain
Other types of bad ads enter the programmatic market in different ways. An inappropriate or off-brand ad can reach a publisher’s page through overly broad targeting or content categorization. Many ads would actually pass a cursory scan or even manual review, even if they would be considered inappropriate for most premium publishers — for example, MLMs, or sites that produce low-quality content and exist only to host ads. Sometimes a bad ad will come from an entirely reputable advertiser, but an advertiser whose message is inappropriate for certain types of content or audiences.
Bad Ad Delivery Timeline
The State of Programmatic Inventory
Managing Demand Partners & Ad Quality
Bads ads volume
39% of publishers say the volume of bad ads coming through the header has increased in the past year.
Bad ads common types
The most common type of bad ad coming through the header are ads with poor or off-spec creative (52% of respondents). Inappropriate formats (in-banner video, pop-ups, etc.) are most common for 10% of respondents.
Deceptive ads
89% of publishers say they see deceptive ads on a weekly basis. GeoEdge research shows over 50% of the malicious ads in the ecosystem are ads with deceptive content.
As mature as the header seems, it is already being shaken up by the seismic changes rocking digital advertising. Check out the chief findings in the GeoEdge Ad Monsters Header Bidding Ebook.
Risks & Dangers: The Cost of Bad Ads
When bad ads reach users, publishers pay for it in the short term by pulling their ad ops and development professionals away from the advanced, specialized work that can grow the business meaningfully. Instead, those professionals are tasked with clean-up work and troubleshooting to find the source of those ads. In the long term, bad ads can pull a publisher’s business into a downward spiral — starting with loss of reputation and trust. Brand safety is high in advertisers’ minds today — as audience targeting became more accurate, advertisers came to focus also on the environments in which their ads appeared. Advertisers have no trouble pulling spending from publishers who don’t meet their brand safety standards. And when trustworthy advertisers walk away, they open the door for unfamiliar advertisers — and for overall yield and CPMs to suffer. Brand safety issues for advertisers lead to revenue issues for publishers.
Brand Safety Challenges for Platforms & Publishers
Of course, brand safety is an issue for publishers as well — users care deeply about the environment content appears in. Bad ads lead users to believe the publisher doesn’t value their engagement (and that the publisher gets paid by untrustworthy sources). And users know that if one of their preferred sites seems to have gone downhill, they’ll be able to get similar content from elsewhere. Ad platforms are aware of how bad ads negatively affect their reputation in the publisher community for trustworthiness and consistency — and how having reliably clean monetization can be a competitive differentiator. Here are some points to illustrate what’s at stake:
Blocking Bad Ads
As important as it is for a publisher’s business that they stop bad ads before they reach the page, many publishers still rely heavily on outdated or inefficient methods. Creative scanning along the supply chain catches some inappropriate ads, but scanning inspects only a percentage of ads coming through — and it’s ineffective against cloaking attacks. Even though tools, such as GeoEdge’s, are available to automate the QA process and block bad ads from the page in real time, many publishers rely on reactive and manual processes: 36% choose manual creative inspection as their top method for handling the problem, and 18% choose removing bad ads after they render as their top method.
Minimizing False Positives
Taking Control of Bad Ads
Detecting Bad Ads
Early on, publisher-serving ad quality vendors used scanning tools to detect suspicious code in ad creative. Though, creative scanning looks at only select samples of the ads served, and bad actors developed cloaking techniques to evade detection even in a scanned sample. To safeguard the ad units on the publisher page itself, creative wrapping tools were developed. The ad quality vendor’s code goes into a wrapper around each ad unit. But this could be an inefficient method, as ad ops and development teams would need to re-integrate the creative wrappers following a page redesign or when adding a new ad unit to the page. Page-level protection unburdens publishers’ internal teams by placing the ad quality vendor’s on the page. This comprehensively protects all ad units on the page, even when page elements are rearranged.
Ad quality tools have also evolved: At one point, threats were flagged by matching malicious code in an ad creative and solely matching it against a list of code used in known threats. Those lists easily became unwieldy as malvertisers developed variable code that couldn’t be matched to previously flagged code. Leading ad quality vendors responded by removing or decoding encryption in bad ads, and by using insights from known threats to detect code that behaves similarly.
Eliminating Bad Ads
GeoEdge takes these advanced methods several steps further by also monitoring ad landing pages — where the greatest threats to users’ security and overall experience often lie — and by blocking potential quality threats in real time, so the publisher may review them for approval or rejection. This gives the publisher greater control over their sites and their ability to monetize.
Stopping bad ads before they become a brand safety, reputational and revenue problem requires cutting-edge technology, stakeholder participation — and a clear ad content strategy customized to the particular needs of your publisher brand, your advertisers’ goals, and your audience’s needs. Follow these best practices to maintain high ad quality standards:
Develop internal guidelines for ad content. Clarifying what you consider a “bad ad.” Identify stakeholders and specify their roles in communicating these guidelines to internal teams and outside partners.
Go beyond standard keyword/vertical blocking. Understand context and nuance on your pages, and create appropriately nuanced keyword lists. Break broad content verticals into more specific categories to align your content strategy with the right ad content. Work with your demand partners to implement granular content categorization. Apply those standards to both ad and landing page content.
Take advantage of QA automation. Add a human review at the end of the process. To facilitate that human review, AdWatch from GeoEdge gives publishers a comprehensive look at all ads rendered on the site.
Create a system for users to report bad ads to you directly as opposed to via social media.
Ad Quality Assessment
Performance stats
CTR
Brand status on social media
Inappropriate Format Ads
Ad Ops
1
Are your performance stats at least consistent over time, or is performance dropping? Poor UX drives users away from a site, mid-session and into the future.
2
Has your CTR risen rapidly? If so, this could actually be a sign that you’re being hit with undesirable clickbait-style ads.
3
Have negative mentions of your brand on social media increased? A growing number of complaints could indicate you’re experiencing sustained campaigns of bad ads.
4
Are your users seeing ads in inappropriate formats (e.g., pop-ups, BV)? This could indicate one of your demand partners has been infiltrated by low-quality advertisers.
5
Are your ad ops weighed down putting out fires to meet their goals? Too much time troubleshooting bad ads indicates you need more advanced protection.
Eliminating Bad Ads: GeoEdge Protection
While vetting ad quality partners, a publisher or an ad platform must look for comprehensive protection against security threats and the whole spectrum of ad content that is inappropriate for their own brand, advertisers and users. New vendors frequently appear on the market, promising to eradicate the current trend in ad security threats.
GeoEdge’s tools, by contrast, can detect and block malvertising as well as deceptive ad content, off-brand ads (per the publisher’s unique standards), adult ad content, prohibited industry verticals, and off-spec ads.
The GeoEdge team brings years of experience in ad quality and security, a proven record of detecting and blocking new quality threats as they emerge, insights from publishers and other digital stakeholders around the world, and a commitment to transparency and communication. Turn to GeoEdge for:
Comprehensive ad security and quality solutions from one trusted partner.