At Possible, generative AI shifts from shiny object to useful tool, but with a long way to go

Google’s announcement last week that it would continue to permit third-party cookies means Chrome is an outlier compared to Apple’s Safari and Mozilla’s Firefox. And, of course, this is not to mention the (already) cookie-less CTV or mobile device landscapes.

Ravi Patel, CEO of SWYM, a startup that uses AI to ease the challenges posed by such fragmentation, told Digiday how many advertisers are exploring AI to map and segment.

“A lot of brands are now starting to think, ‘How do we have smaller portions of audiences and use more AI with better methodologies to optimize the actual outcomes you’re looking for, as opposed to buying audiences blindly,” Patel explained.

None of this should be surprising to anyone who glanced at the conference agenda. AI was everywhere — whether front and center on panel titles or bubbling up naturally in conversations across the main stage.

Take Monday’s panel led by Tyler Romasco, svp of global publisher development at ad tech vendor OpenX. The discussion, with execs from CVS Media Exchange, ANA, and TAG, centered on inventory quality but AI still made its way in. It came up as a tool to help tackle fraud, detect bot traffic and flag harmful content. Less shiny object, more practical utility. 

“What we saw at Possible this week was that overwhelmingly, people are thinking about using AI two different ways,” said Romasco. “The first is efficiency and operations. I heard a lot of conversations around leveraging AI to make businesses more efficient and ultimately service customers better. People are excited about the opportunity this can open up to spend time on strategic innovations.”

To Jon Halvorson, svp of global consumer experience at Mondelez, the value of generative AI lies in content and conversion. “We care about how we’ll use AI to improve the quality of our content, and we care about how we’re gonna use AI to improve our e-commerce operations,” he told Digiday at the Digiday Studio at Possible (Digiday is a media partner of the conference). “We can take our content to the next level by ultimately improving the quality, taking the cost of these incremental assets down to zero. And in AI for commerce … you look at even some of our storied brands, Ritz and Oreo — we don’t convert 50% of the people who get our product pages into a sale. What a huge upside for the business.”

It’s a reflection of a broader shift across advertising: the promise of AI to boost performance and sharpen programmatic strategies is real — but so is the need for more transparency around how the tech works and why it’s being used.

“The conversation [at Possible] remained grounded in more immediate concerns like platform volatility, data privacy, measurement, curation and keeping up with ever-evolving creator strategies,” said Charlie Johnson, vp of international at IP intelligence company Digital Envoy. “The excitement about AI is real, but so is the uncertainty, and for now, most people are watching and waiting rather than diving in headfirst.”

That tracks to some degree. For all the noise, marketers are still early in the AI era — closer to base camp than the summit. Which is why so many discussions at the conference skimmed the surface. Not because the topic lacks depth but because the industry isn’t quite ready to go there. There’s still a long road ahead before AI starts delivering on the bold claims clogging up LinkedIn feeds.

“There was interest in understanding how each AI use case can drive real value for marketers and agencies, whether that value is time saved, performance results, or lift, said Jason Downie, global chief revenue office of RAISING — the tech division of marketing and tech consulting firm Making Science. “One example of this was looking at how hyper-personalization of creative leads to better performance in search, programmatic, and so on.”

To the degree it’s been developed, agentic AI is also finding its footing among marketers and tech firms. Rob Emrich, chairman and founder of Infillion, explained agentic’s adoption can help with workflow issues around a tech firm that’s offering multiple products. “The first step for a company like us is to make all of the different services, all of the underlying products — whether it’s our creative product, whether it’s a targeting product, whether it’s our the logic in our bidding — that each one of them is separable and is usable by an agent,” said Emrich. “We are in the middle of that process now of essentially upgrading the protocol that our APIs use in order to make them usable by more complex agentic AI.”

Creative was another recurring flashpoint. But as Matt Barash, chief commercial officer at AI ad platform Nova noted, a lot of the AI talk still came wrapped in buzzwords and overpromises. 

“The buzz around the proliferation of AI over the past few months has catapulted the opportunity around creative to the top of the industry agenda,” said Barash. “Far too many attendees were trying to appear forward thinking and lean into their AI offerings but often mistakenly — when it really means automation or machine learning. The devil is in the details and more often than not the conversations were largely surface level.”

In that sense, Possible became less of a hypefest and more of a reality check: who’s actually ready to put AI to work and who’s still stuck in pitch deck mode?

“Unlike the last few years at these events, AI has felt like a buzzword latched onto all conversations without people really having an idea of its potential,” said Wilfried Schobeiri, chief technology officer at ad tech vendor Ogury. “The dialogue now definitely seems more organic. The industry isn’t just talking about AI anymore, it’s actively utilizing it and has a clear sense of what it can do — especially from a productivity perspective.”