There’s much more to advertising quality than glossy creative, celebrity models and catchy slogans. These might be all that register with audiences, but between concept and delivery are a huge number of creative and strategic decisions and a web of technologies that work to keep the ad quality bar high on the open web. We may not notice ad quality when everything’s working, but we definitely notice when it isn’t.

What Determines Ad Quality? Everything

From design to delivery, every step of the ad journey contributes to its quality. At their core, ads should be visually appealing, conceptually strong and aligned with the brand’s message. The creative assets and the ad wrapper should be lightweight and not overladen with third-party tags, as a seamless and performant ad experience enhances user engagement and fosters a positive perception of the product or service.

Beyond aesthetic and technical considerations, transparency and authenticity are paramount for ensuring customer safety in advertising. Any sponsored content must be clearly labeled (this is particularly important in native campaigns where ads might not be immediately apparent), and the user must be guided to the expected product or service should they click. Misleading or deceptive ads not only tarnish a brand’s reputation but also lower the user’s estimation of advertising in general.

However, context is key, and all the work put into creating a high-quality ad means nothing if it’s delivered to an unsuitable environment. Should an ad land on a website where users are frustrated by poor site performance and overwhelmed with banners and pop-ups, any resulting views or clicks are more likely to be accidents than signs of interest or intent.

Advertisers who care about quality should ensure this quality is reflected on the sell side, whether through private marketplaces, direct deals or publisher networks. Content farms that churn out search engine-optimized pages might deliver traffic, but this traffic is made up of one-off visitors, not loyal audiences that will be more responsive to advertising because they trust in the publisher, who in turn can transform these loyal audiences into targetable segments.

Poor placements both undermine performance and can also pose a brand safety issue. Should brands appear alongside misleading or even harmful content, there is the risk of reputational damage both through the brand’s adjacency to the content and the implication that the brand is endorsing it with its ad dollars.

Shielding Users From Bad Ads

It’s not just up to brands, agencies and publishers to ensure that users experience high-quality ads. The networks that take ads from A to B also need to moderate the supply chain to prevent low-quality or harmful ads from making it to the other side. Filtering out ads that would otherwise clog the internet with trash requires a mix of automated and human moderation.

This multifactor moderation begins with machine learning models that apply vast swathes of historical data to predict which creative will not hit the quality bar. Such models should encompass every component of the creative to minimize the volume of false flags and provide detailed reporting so that manual moderators can investigate ambiguous cases and further refine the model’s accuracy over time.

Automated moderation significantly accelerates and simplifies the process, but a human touch is always required to pick up on subtle contextual cues and sensitivities that users may have depending on their demographics. This is also why ads should always provide the option for users to report them. A machine learning model might be able to recognize clickbait, but it won’t necessarily recognize when a creative might cause offense or is misaligned with the publisher’s values.

Moderation teams must be available locally in each region where an advertising network is present, as much of what makes an ad inappropriate might be highly specific to a particular culture or group, and can change rapidly depending on recent events. Where possible, the diversity in moderation teams should reflect the diversity of real-world audiences.

Will AI Be A Blessing Or A Curse For Ad Quality?

Outside of bad actors, much of what determines ad quality is simply a matter of resources. AI tools can help raise the bar for ad quality across the board by making asset creation, formatting and optimization widely accessible, along with the ability to automatically align advertising with user signals such as location, time and language, as well as audience segment.

AI also has the potential to solve the long-standing issue of blocklist-based brand safety tools preventing ads from showing up in premium environments due to overzealous blocking. Contextually aware AI engines that can understand the distinction between shooting a movie and shooting a gun mean that ads from brands with a higher safety threshold — which typically also have a higher quality bar — can appear in environments where they previously would have been blocked.

Of course, the needle can swing the other way too, with the ease of asset production leading to a flood of low-quality AI-generated ads, including from bad actors. Anyone who has spent time on X (formerly Twitter) lately will likely have noticed bizarre AI-generated ads that don’t even mention a product or are riddled with visual and textual errors, with many deliberately masking that they are ads or, worse, blatantly fraudulent.

Tight moderation will be key to preventing such ads from spreading to other platforms and networks. This has become a matter of AI versus AI, with AI-powered moderation used to detect and block low-quality AI-generated ads.

From advertisers to publishers and the technology partners in between, everyone involved in the digital advertising supply chain is responsible for enshrining a high-quality ad experience. By leveraging AI-powered technical solutions and human moderation, advertisers who prioritize quality and integrity will be rewarded with access to audiences, who can navigate the web without worrying about irritating, offensive or fraudulent ads. Protecting ad quality at this time of rapid technological change is in everyone’s interest.

(As published on Forbes)