permanent ad that becomes irrelevant after the first listen. Spontaneous creation and automation Artificial intelligence has certainly made its mark on the entertainment industries, and now it’s coming for ad tech, too. “AI is both solving traditional ad-tech problems and changing media company workflows, leading many people to reassess their existing partnerships and advertising stacks,” shares Shaw. “A new wave of AI companies, some highly specialised, others more general, are entering spaces that ad-tech companies once dominated, integrating focused AI capabilities into established platforms.” Besides catalysing mergers and acquisitions and generally reshuffling the corporate landscape, AI is expanding what ads can do – and how quickly they can take shape. In particular, large language models (LLMs) and generative AI are leading the charge in letting advertisers make live adjustments. “Imagine the ability to generate copy, adjust visual layouts or even synthesise personalised voiceover in real time based on audience signals,” Lods suggests. “The ability to produce thousands of creative variants at low cost changes the economics of personalised advertising on a fundamental level.” Artificial intelligence, while useful in efficient ad creation, also touches video content analysis, ‘dramatically improving’ this process, according to Lods. “Contextual targeting – targeting based on what content surrounds an ad rather than who the viewer is – has become a much more viable, privacy-safe alternative to audience-based targeting, in part because LLMs understand nuance, sentiment and brand safety risk in ways that earlier keyword-based systems were simply not able to,” he explains. Lods is right to mention privacy. Local regulations and the availability of information pose barriers to genuine personalisation, thus limiting an ad’s effectiveness. While companies might argue that compliance hinders business,
consumers do – presently, at least – have the right to keep some of their information (demographics, viewing habits and the like) to themselves. Failings and future opportunities In the simplest terms, ad tech exists to turn audience engagement into earnings, but sometimes that return on investment is tricky to quantify. To account for the dynamism of digital environments, Shaw says: “Advertisers are moving from traditional reach-and-frequency models and longitudinal brand studies to real-time, user-level measurements. They can now link exposures to actions – like website visits, purchases or app interactions – using granular data. “Advanced measurement is the next frontier,” he adds. “As the industry moves away from panel- based audience tracking to hyper- personalised, campaign-centric metrics, there is a big opportunity to bridge gaps and improve insights for advertisers while enhancing the viewer experience.” For now, faulty measurement tactics and other key issues, such as fragmentation, remain. “Inconsistent standards across publishers can lead to overexposure,” Shaw explains. “Viewers may see the same ad ten to 40 times, even within the same show.” If the goal is engagement, that’s a sure-fire way to fail. “While ad tech can effectively emulate the traditional TV viewing experience, there are many areas where it can improve,” adds Lods, who points out similar shortcomings. “Challenges remain when trying to maintain viewer experience at scale, especially when ad loads or latency are poorly managed.” For Lods, “the future lies in lower-latency ad insertion for live content, improved standards such as SGAI, as well as shoppable and interactive ad formats.” But, in the long term, he sees ad tech moving towards real-time, personalised creative assembly, or put in simpler terms: “ads built dynamically for the individual in the moment. This is
where generative AI and DAI infrastructure are heading.”
Powered by FlippingBook