by Csaba Szabo, EMEA Managing Director, Integral Ad Science (IAS)

Cookies/Contextual

There are three major regulatory influences on the digital advertising programmatic ecosystem right now:

1) Privacy changes – Laws and regulations such as the GPDR and CCPA now restrict the use of audience insight without explicit consent, driven by a demand from consumers in protecting their digital privacy.

2) Device level – Apple IDFA will restrict the use of advertising identifiers.

3) Browser and app level – Google Chrome is the latest browser to announce it will not support third-party cookies as of 2022, following in the footsteps of both Firefox and Safari.

All three of these shifts essentially contribute to the same net effect: restricting advertisers in reaching specific audiences. They make audience targeting less reliable for marketers who are now turning to alternative strategies to ensure their campaigns are impactful. The industry is innovating to find an alternative way to buy media campaigns after an over-reliance on audience and behavioural data. The discussion around finding a ‘future proof’ solution has circled around balancing privacy protection with the need for ROI. 

Marketers will always need to protect their brand message and reputation online. To add to the complexities, every brand has its own perspective, requiring a different approach on where they should or should not appear, based on their unique brand values. 

At IAS, we believe that contextual targeting is widely regarded as one of the most effective solutions available and advertisers are leaning into context given the imminent challenges the market is facing.

It’s important to recognize that brands, agencies, and publishers are all benefiting from the rise of scaled contextual technology. For advertisers, understanding context helps them build more impactful connections with their audiences by finding the optimal adjacencies for their campaign messaging. For publishers, smarter and more granular content classification means that they can package their inventory to be more relevant for both advertisers and consumers; this in-turn unlocks extra yield.

What’s also clear from the past year is just how critical the quality of advertising media is for the success of the entire ecosystem. Not all ad impressions are equal and everyone wins when campaigns find the best quality impressions for their objectives.

Artificial Intelligence 

The world of AI automation is one to watch. With momentum building over the last few years, we expect this growth to accelerate in 2022. Digital advertising has made a lot of exciting progress in automation, most notably with programmatic marketplaces, helping many different companies in the value chain to process, enrich, and validate larger volumes of data at scale never seen before.

IAS has made significant strides in AI automation, and we see more opportunities ahead! We build, test, deploy, and monitor in production hundreds of machine learning models — the core of AI. We constantly refine and develop our infrastructure to support new tools; for example, adding automatic capabilities to explain why models make the predictions they make.

For example, the digital advertising landscape is changing rapidly, and contextual targeting is now fundamental for the proper optimisation of campaigns. Contextual advertising has evolved into a more effective way for brands to capture the emotion of their audience. Advancements in AI and cognitive technologies, such as natural language processing, can now determine the sentiment conveyed on a given page at scale, enabling more suitable ad placement decisions.

In particular, these technologies analyse the full text and the relationships words have with one another in order to accurately comprehend the context of the page. The overall sentiment (positive, neutral, or negative) is understood along with the associated emotional classifications. It also offers increased capabilities for localisation, translating copy into different languages without losing or detracting from the sentiment of the ad campaign.

It’s one thing to say a piece of digital content is positive but determining whether it’s focused on love or humour is something else. There are now smart tools that have developed huge semantic networks covering all 500,000 concepts of language and the relationships between words, in addition to diverse emotional contexts linked to more than 60 grades of human emotion.The combination of AI, machine learning and sophisticated technologies – natural language understanding and semantic analysis – allows for human-like comprehension: such as the capacity to immediately spot the difference made by small linguistic variations, from changes in tense right down to the use of a full stop.

What all of this delivers for brands is comprehensive real-time intelligence that enables precise ad matching for multiple contextual factors including emotion, at scale. With AI technologies being sophisticated enough to control how a message is presented in various contexts, advertisers now have far greater agility when it comes to planning and executing their campaigns.

Metaverse

With the early hype surrounding the metaverse turning into sustained excitement, it is unsurprising that projections say that it could be worth $800 billion by 2024.

For marketers, the metaverse will throw up issues that will seem familiar – measurement, brand safety – in an environment that has many new content forms. While speculating exactly what shape the metaverse will take is tough, what is clear is that marketers will need to harness the lessons learnt in web2 to tackle new issues in web3.

The dawn of the metaverse brings with it new content forms for marketers and brands to wrap their heads around. The vision of the metaverse as a fully immersive world accessible by all may be some way off, but AR and VR-enabled experiences are already impressing. In fact, 3D measurement capabilities for in-game advertising may be the first hurdle for marketers, as it seems the more natural stepping stone into a fully immersive experience, and gaming is undoubtedly a big driver for wider metaverse adoption. In essence, enabling the capabilities for current gaming environments will still be relevant in virtual ones down the road.

While user intent now may be tracked by clicks and mouse tracking, the metaverse will be able to take into account a whole range of other signals. For example, a marketer may be able to measure a consumers’ body movement in the metaverse to gauge engagement and interest in a piece of content.

But as brands navigate this new ecosystem, they will need to remain diligent. Where consumers go, bad actors usually follow. As the industry saw with the burgeoning of the video content market just over a decade ago, ad fraud is a serious risk, leading to overinflated stats and wasted budget.

To stop these same issues repeating in web3, industry-wide quality and measurement standards should be built into the very foundation of the metaverse. It is imperative that we learn the lessons from the previous iteration of the internet and implement measures to make new digital environments safer and more secure for both brands and users from the outset.

But ad fraud is not the only problem likely to migrate from web2 to web3. The fast-moving, interactive, hyper-visual new environment of the metaverse will create further safety issues and brands may run the risk of appearing next to inappropriate content. On top of this, consumer concerns over privacy will likely grow leading to ever-tightening privacy legislation.

Brand safety and suitability solutions can be the answer to both these problems. This powerful targeting method not only keeps brands safe, but provides a cost efficient way to reach target audiences, by placing advertisements in environments that are deemed both safe and contextually relevant by advanced AI.

Ultimately, it’s still early days for a fully fledged digital advertising ecosystem in the metaverse. We’re still at the ‘test and learn’ phase, but it’s important to get a head start. Undoubtedly, if it’s where users go, marketers will follow; and the need for verification and digital media quality will not be far behind.

Making the metaverse a safer place for both brands and users will take more than tech solutions. The advertisers will need to collaborate in order to create standardised guidelines and a shared language in which to talk about brand safety and suitability in this brave new world – much like what GARM is doing now.

The metaverse offers endless possibilities for brands. But in order to take full advantage, we must take heed learnings of the past and build a new environment that all are able to enjoy.

Brand Suitability

Brands invest a significant amount of time creating an image, cultivating consumer perception and fostering long-term associations. Therefore, it’s important to ensure that digital messages appear in safe and suitable environments. This not only reduces risk, but also ensures that the right consumers are reached.

Last year saw an industry evolution from brand safety to brand suitability. Born from the need for more nuanced brand stewardship, this move was accelerated by a massive uptake of digital engagement from consumers and a shifting news agenda that made it even more important for each brand to direct spend towards the most suitable environments.

Where traditional brand safety tools relied on blanket approaches to avoid unsafe or inappropriate content, brand suitability is more nuanced and attuned to individual brand risk sensitivities.

Advertisers need to carefully assess what their brand safety and suitability thresholds are and ensure that the right tools are in place such that their ads aren’t placed in any unsuitable environments.

Where blocking tactics and keyword exclusion lists rule out any content featuring broad and often outdated terms, such as crisis events, suitability looks deeper. Once brands have set their safety requirements – threats involving specific types of content covered by standard categories and execution lists – they can define their acceptable level of risk for certain topics and terms and use custom analysis to assess every placement.

For programmatic buys, using pre-bid filters ensures that brands only pay for quality impressions that are suitable. Buyers then do not bid on impressions that will be wasted due to brand safety concerns. When pre-bid filters are applied, analysis occurs in the DSP ahead of the bid and prevents brands from buying on unsuitable impressions. Finally, pre-bid filters allow savings on media cost and data fees associated with those unsuitable impressions

Navigating brand risk is getting better with technology. It’s true that the lockdown fueled digital consumption and higher content production than ever, at both the user-generated and publisher level. This increased volume makes ad misplacement more likely. But thanks to the emergence of sophisticated and purpose-built analytical technologies, advertisers now have tools to instantly gain precise ad targeting techniques on a large scale.

Programmatic

Programmatic advertising is changing rapidly, and contextual targeting is now fundamental for the proper optimisation of campaigns. Key to delivering ads that are contextually relevant is the ability to understand consumer sentiment and emotion.

Sentiment and emotion is nothing new in advertising. The advertising industry has a rich history of harnessing feelings to forge a connection: brands have long aimed to go beyond simply educating and persuading consumers by tapping emotional triggers that influence their decisions. To secure maximum outcomes and brand suitability, striving for relevancy is the broad goal.

Advancements in machine learning and cognitive technologies, such as natural language processing, can now determine the sentiment conveyed on a given page at scale, enabling more suitable ad placement decisions. In particular, these technologies analyse the full text and the relationships words have with one another in order to accurately comprehend the context of the page. The overall sentiment (positive, neutral, or negative) is understood along with the associated emotional classifications. For example, amusement, love, and hope are different emotions that can be classified within positive sentiment, even if those specific words are absent. In order to understand sentiment and emotions elicited by the text a probabilistic approach using keywords in isolation will never be accurate.

Due to its efficiency, data, and scalability, programmatic advertising is not going away, and the use of technology to understand suitable, quality ad placements will increase. This is what IAS defines as a Quality Impression™- an ad that is viewable, by a real person, in a safe and suitable environment, and all within the correct geography.