The past couple of years have seen an increased focus in the industry on ad specs and behavior, in particular from organizations like the IAB and Big Tech businesses like Google. This is part of the ongoing fight for ad quality, taking place on three fronts. Security, ad content (including inappropriate, off-brand, irrelevant or otherwise undesirable content), and ad specs/behavior. When it comes to heavy ad monitoring and prevention, the consequences of inaction can be costly to publishers.
Whether these ads are served in good or bad faith, digital publishers need to take measures to keep them off the page so that users continue monetizing the site and ad partners remain satisfied.
What Is a Heavy Ad?
In 2019, Google set standards for heavy ads: An ad may use no more than 15 seconds of CPU within a 30-second span, for a total of no more than 60 seconds of CPU, and no more than 4MB of network data. If an ad exceeds those limits, and if a user has not interacted with it, Chrome will block the ad and serve an error message in the ad unit. Google has published more granular guidance for developers.
Google set these standards after researching ad performance, discovering that 0.3% of all digital ads are responsible for eating up 27% of network data and 28% of CPU among all ads in the ecosystem. IAB Lean issued similar guidance. Following these guidelines allows pages to load faster, preserves carriers’ network bandwidth, limits users’ data overages, and even discourages crypto-mining attacks.
Heavy Ads Cause Problems for Users
Many ad security threats and low-quality content ads come from overtly malicious actors. However, ads that compromise page performance or the resources of a user’s device are more of a mixed bag. Some are part of crypto mining schemes or arbitrage strategies. Others are simply coded poorly or are not optimized for battery or data usage.
Heavy ads can be insidious because a user won’t necessarily recognize that their poor performance could be anything beyond a mild inconvenience. It’s easy to notice when a page loads slowly. Excessive CPU or data usage is more difficult to detect — at least, at the outset. The impact of heavy ads is often noticeable and includes devices shutting down mid-session, overage charges, a user’s incorrect assumption that the device needs to be serviced, and more.
Why The Industry Needs Heavy Ad Intervention
In 2020, Google introduced Heavy Ad Intervention (HAI) to enforce its heavy ad policies. HAI is a helpful way to ensure a safe browsing experience. It encourages engagement, benefitting publishers and advertisers alike. HAI is necessary because communication breakdowns and security lapses along the ad supply chain are ultimately unavoidable without robust, real-time ad quality tools.
HAI aims to block resource-straining ads such as malicious crypto-mining attacks, improperly compressed images, and in-banner videos. The initiative was based on IAB Lean protocol, along with insights from the heaviest .01% of ads seen on Chrome.
HAI replaces a blocked ad with an error message, saying that the ad has been removed and showing a link the user can follow for details. Error messages disrupt a site’s overall feel and brand, and can lead the user to suspect the publisher has partnered with shady ad partners or simply has low standards for the content they host on the page. Mistrust causes users to avoid a site and prevents the publisher from monetizing the user’s lifetime value.
How HAI Impacts Publishers’ Business
When a heavy ad causes a page to load slowly or freeze, the impact is immediate. Research shows 47% of users will stop engaging with content if they think it’s taking too long to load. On mobile, increasing page load from just one to three seconds makes a user 32% more likely to stop engaging.
Good communication along the supply chain isn’t enough to keep all heavy ads off a page. Think back to when Flash was being phased out — for years, publishers continued getting Flash ad content, regardless of publisher specs and industry guidance.
How Do I Know What HAI Is Blocking on My Site?
There are preventive measures publishers can take, though they are not all equally reliable. Google has put forth resources for developers to test HAI on their sites. A publisher’s team can use it to select ad content and see any “ad removed” messages HAI would have displayed.
HAI launched before the programmatic ecosystem was equipped to report on it. According to industry insiders, as of the summer of 2021, it appears that no ad servers can report on HAI-blocked ads. Google’s advice to publishers is that they don’t need to take special measures with ads from third-party sources — as long as they bear in mind that heavy ads will trigger an “ad removed” error message.
There’s no consensus on the share of heavy ads blocked because of data and network bandwidth, versus CPU drain. There’s also no agreement as to whether heavy ads are more likely to come from programmatic or direct advertising. Moreover, cryptomining attacks will often come through broad ad content verticals. GeoEdge has found that the most egregious heavy ads, including ads that use up to 75% of CPU, come through the “internet” content category. With spot-checking being such an ineffective strategy, publishers must consider ad quality tools specifically designed to protect all of their inventory from disruptive and revenue-threatening heavy ads.
Navigating HAI Reporting
Publishers should always communicate their own specs, along with broader industry standards, to their demand partners. And, demand partners should always ensure all Javascript is well-written and complies with Google’s guidelines for animations. Publishers should, however, expect that heavy ads will come along, and that they’ll be blocked, one way or another.
Publishers need to take control over their inventory and reporting to prevent user experience and monetization from being compromised by heavy ads and error messages. They need an ad quality solution that is efficient and consistently reliable.
