The fall of the cookie, the rise of brain activity-based marketing, new challenges in tracking multicultural data, reinvigorated debate on “fair use” in training AI, and innovation in gen AI use cases
As we march towards the end of 2024, we’re continuing to bring you the most notable trends from some of the top outlets and writers covering all things ad and martech. This month’s key themes include: innovative new uses cases for AI; evolving legal standards with multicultural data and fair use of copyrighted content; scientifically-backed, emotion-based targeting; and the continued industry shift away from the cookie.
With all of the back-and-forth around the deprecation of third-party cookies, marketers are reaching the point of exhaustion. While a small-scale cookie-buying marketplace still exists, engineering teams and budgets have begun to de-prioritize cookies and invest alternatives that have staying power and are future-proofed.
As AdExchanger noted, “Google introduced more uncertainty when it proclaimed that it would drop the cookie. Markets don’t like uncertainty. Hundreds of companies started to slowly pivot. As each deadline approached, the pivots sped up. Today, no company is putting effort toward the cookie. Uncertainty prevailed in doing what overt declarations could not. Our ecosystem has moved on.”
As a result, the advertising and martech industries are pivoting, prioritizing transparency and user privacy. One emerging solution for this is alternative IDs, technologies or methods used to track users’ behavior across websites and apps that do not use cookies. Generally speaking, alternative IDs fall into two categories: deterministic and probabilistic. Deterministic IDs as those that rely on personally identifiable information (PII), such as an email address, that is then concealed to protect the user’s identity. Probabilistic IDs, as the name suggests, look to a variety of signals such as IP address, device type, screen resolution, and operating system to gather a best guess as to who the user is. Probabilistic IDs also do not rely on any first-party data being collected.
However, in order for these alternative ID providers to avoid the same fate as the cookie, they must become advocates for user privacy and audiences must become educated in the choices they have for having their data tracked. Audiences should understand the value that is being provided to them by opting in to these alternative IDs. Furthermore, alternative ID providers must demonstrate value to the platforms seeking to use them, be efficient and inexpensive, in order to ensure their success.
No surprise here—the widening scope of uses for gen AI continues to grow. This month, MediaPost covered Amazon’s announcement of its new application suite, AI Creative Studio, where marketers can utilize AI-powered ad generators for images, video and audio, allowing “brands to conceptualize, create, and refresh content regardless of format in one application.”
One of the tools included in the suite is a generative AI to create audio ads that can run across Alexa-enabled devices. Once the AI is given the product for the ad, it will analyze the product details and customer reviews to generate copy. Then the advertiser will select the synthetic voice and its tone, along with any background music that suits the ad. When put together, this tool can easily generate hundreds of variations of an audio ad, personalizing it for a wide variety of audiences.
This tool also promises the ability to convert ads from one format to another across platforms. The suite will additionally provide metrics such as click-through rates in platform, making it easy to track performance.
Another, perhaps more unexpected, use case for gen AI is encouraging people to rethink, and potentially change, conspiracy theory related beliefs. In a recent article from Fast Company, researchers from MIT and Cornell conducted a study where conspiracy theorists told a chatbot about a theory they believe in and their evidence for supporting it. The chatbot, being trained in responding accurately to the false information relayed to it, then conversed with the subjects on the topic. The study demonstrated that 20% of participants changed their beliefs after having the conversation with the chatbot.
The article goes on to unpack why the chatbot was effective, “Conversation with AI creates a healthy dissociation from another human being. I suspect that separation is what enabled the subjects to rethink their feelings. It gave them emotional space. They did not become defensive because their feelings were not hurt, nor their intelligence demeaned. That was all washed away, so the subjects were able to actually ‘hear’ the data—to let it in to trigger reconsideration.” It’s exciting to consider that AI has promise in the effort to dismantle misinformation and promote a truth-grounded society.
As privacy continues to grow in importance, some states have recently begun to introduce legislation that protects race and ethnicity, classifying it as sensitive, personal information. In an article from Ad Age, this may be an unintended double-edged sword: “privacy laws cover race and ethnicity data because of its potential use to discriminate. But not having it available is likely to result in multicultural audiences being undercounted in traditional measures or missed by targeted advertising.”
While these laws were arguably derived from a good place, the loss of multicultural data poses a problem for both marketers and consumers. Carlos Santiago, co-founder of the Association of National Advertisers’ Alliance for Inclusive and Multicultural Marketing (AIMM) notes, “losing access to such data also will make it dramatically more difficult for marketers to fulfill their responsibilities, including building representative audiences, designing inclusive marketing efforts, ensuring survey samples and panels are representative and measuring the effectiveness of their campaigns on key audience segments.” In order to still reach multicultural audiences, companies are creating representative models using racial and ethnic data that has been provided ethically and consensually.
Beyond certain states beginning to include race and ethnicity in their privacy laws, the Federal Trade Commission (FTC) has also begun to treat racial and ethnic identification as sensitive personal information in some recent decisions, signaling that it is increasingly becoming closer to the General Data Protection Regulation (GDPR) in terms of privacy regulations.
Navigating these privacy-related regulations is also becoming increasingly complicated from a litigation standpoint, as industry data use and law mandates are often out of lockstep. In Europe alone, a 50-billion-euro fund has been established for suing ad-tech companies that violate data privacy policy.
Following the ever-changing privacy policy landscape while maintaining a representative audience for ads will certainly be a challenge, but a worthwhile one.
Another growing interest this month in the world of data-driven advertising is the connection between emotions, brain activity, and ads. Recently, Ad Age reported on a new marketing strategy launched by TopGolf and its partners that used data collected from their customers’ brain activity to inform its advertising. The goal of this new strategy was to replicate the joy people feel while playing Top Golf in the ad, and see if that impacted consumer engagement.
“With a lot of digital campaigns, you’re optimizing and placing media based on digital metrics [such as click-through and impressions], and you’re implying what the person actually felt based on that digital engagement,” said Laurel Boyd, chief creative media officer, Mediahub, one of Top Golf’s partners. Boyd said this new strategy “was a way to really get at the root with empirical data to know how people were feeling.”
The results of the partnership provided fascinating insights, demonstrating that ads that evoke the same feeling customers have while playing TopGolf result in better engagement and brand favorability. In comparison to previous campaigns, the ads that aligned with this emotion-based strategy saw a 193% increase in engagement and 13% lift in brand favorability. Clearly, emotions play a huge role in advertising and this new research is demonstrating empirically just how effective emotion-based ad targeting can be.
In order for gen AI to perform in the wide variety of ways it can today, it must be trained on massive amounts of data. Sometimes, included in the data used to train these models is copyrighted content, sparking a new debate this month on the expanding definition of “fair use”.
Forbes recently explored the way AI analyzes and uses copyrighted content to delve further into potential legal hurdles. “For an AI model to learn what a dog looks like, it needs to analyze millions of dog photos. The system isn’t interested in the artistic composition or the specific dog in each photo – elements that might be protected by copyright. Instead, it’s learning to recognize general features like fur, four legs, tails, and typical dog shapes. In fact, ‘verbatim copying’ is the necessary intermediate step toward accessing the unprotectable ‘ideas and functional elements’ of works that allow AI systems to learn generalizable patterns and concepts rather than simply memorizing specific content. AI models instead encode patterns from training data into parameters, generating responses using learned probabilities and not by referencing stored content.”
As such, since the AI is not purely copying the “ideas and functional elements,” using copyrighted content to train an AI can be considered “fair use,” despite certain copyright owners taking legal action against AI companies for using their content to train AI without their consent. Additionally, legal scholars note: “The input data used for AI training may often be permitted under ‘fair use’ or ‘fair learning.’ At the same time, purely machine-produced output is typically not copyrightable.”
If copyrighted works were to be restricted for training AI, a concern is that AI will train itself on less verified content and more AI-generated content. “Restricting AI’s access to the vast repository of data and knowledge, on the other hand, will severely hamper its ability to learn, significantly impeding potential advancements and curtailing broad societal benefits. And, if AI trains and learns from its own outputs or synthetic data, it risks a downward spiral where its output becomes less varied, less accurate, and of lower quality, leading to hallucinations—a process researchers call ‘model collapse.’”
Ultimately, at the center of this debate is how fair use is meant to protect the ability to build upon ideas. The growth process is made possible by gathering and analyzing information that has already been published. This is true whether a human or machine is doing the learning.
Navigating the future of B2B tech marketing: Enhancing data quality and processes (MarTech Series)
The marketing landscape is undergoing a seismic shift and B2B tech marketers must stay nimble to keep pace with change. The phase-out of third-party cookies, the introduction of IP address masking, and increasingly complex privacy laws are fundamentally reshaping how data is collected and used. In this new environment, data quality has become the cornerstone of effective marketing strategies. To gain sharper insights, improve targeting, and maximize ROI, marketers need to focus on building accurate, reliable, and unified data processes. This not only strengthens current marketing efforts but also future-proofs brands against the evolving challenges of online and offline marketing. Marc Sabatini, SVP of Enterprise Solutions
17 big tech trends agency pros are excited about (and why) (Forbes)
Conversational user interfaces are developing rapidly. Consumers expect brands to predict their needs and deliver personalized experiences. Advances in natural language processing, machine learning and artificial intelligence, combined with better processing and connectivity, make CUIs more intuitive by anticipating needs based on past behavior and real-time context, improving sales, marketing and customer service while disrupting digital advertising. – Thomas Zawacki, President, Axle Agency / Chief Marketing Officer
The future of identity resolution: You have the power (The NonProfit Times)
As the digital landscape continues to evolve, the way nonprofits engage with their donors is fundamentally changing. At the heart of this transformation lies identity resolution—a powerful tool that allows you to create a unified view of your supporters across multiple touchpoints. With the deprecation of third-party cookies, the rise of privacy regulations, and the increasing complexity of donor interactions, the need to effectively harness identity resolution is more critical than ever. Niely Shams – President, Data Axle Nonprofit
Courtney is a seasoned communications and public relations professional with 17+ years of experience working in both the public and private sectors in diverse leadership roles. As Data Axle’s Senior Public Relations Manager, she is intently focused on elevating the company’s media relations presence and increasing brand loyalty and awareness through landing coverage in top-tier media outlets.