Your next customer isn't human, according to industry experts at IAB Australia's Discovery AI and Search Summit today.
Karin Du Chenne, executive group director at Kantar, said the stakes were higher than most brands realised, and that company size offered no protection.
"Just because you're big doesn't mean you're safe," she said.
The brands benefiting from the shift were not necessarily the most recognisable.
Du Chenne said LLMs were surfacing smaller, more specialised competitors above household names simply because their messaging was cleaner and more consistent.
"We see that very specialist brands, or brands that have been targeting very clear, niche, and consistent messaging are more likely to come through than your everyday brands that are targeting the masses."
Most brands, she warned, had no idea how to respond, because the skillset that built them was the wrong one for this audience.
"We've all been pretty good at marketing to people, touching their hearts, touching their minds. But now, how do we market to this new consumer, the machine?"
The reason comes down to how LLMs actually read.
Rather than seeking a single authoritative source, LLMs draw on consistent semantic patterns across multiple sources, which means the warm, emotionally diverse language that resonates with humans is effectively invisible to them.
"Large language models cannot read warm and fluffy adverbs. We have to give them content that they can read and access,” she said.
Du Chenne emphasised understanding how different AI systems pulled content was critical to getting the strategy right.
Google AI overviews favour the most trusted, authoritative and verifiable source, presented in a structured way.
While Generative AI engines, predict the best answer based on how consistently and meaningfully a brand appears across third-party sources.
Because of this, one platform brands were underestimating, she said, was Reddit.
Its community-driven, conversational format making it particularly influential in how LLMs surface answers.
"Reddit is playing an above-average share of impact when it comes to geo because it's a community-based platform. People speak, they share — they speak in normal human language, and that's what the machines love."
For brands looking to check the health of their brand amongst AI models, Du Chenne offered a simple diagnostic.
"If you want to see how you do show up on Reddit, just type into the search bar source and put your website there,” said the Kantar director.
“It'll give you a lovely AI summary of exactly what people are saying about you on Reddit. And if this is not what you want to hear about yourself on Reddit, this is your first point of action."
Despite the profound shift in how machines read content, Adrian Vallelonga, head of performance solutions and retail media at Google, said the industry should not interpret the rise of LLMs as a reason to abandon search engine optimisation in their content strategy shifts.
"Search is not dead. It's probably bigger than it's ever been," said Vallelonga.
Renee Stopps, regional sales director of partner sales APAC at Microsoft, agreed, framing the moment not as a disruption but as the next stage of an ongoing evolution.
"Definitely search is not dead. We're kind of revolutionising what it used to be," she said.
Stopps described the shift through three eras of the web, with the current transition moving from open-tab browsing toward conversational and eventually agentic AI, where the machine doesn't just answer user input questions, but takes action on behalf of the user autonomously.
Though, she warned in the current industry state, the quality of agentic behaviour will only be as good as the input brands feed them.
"The AI output is only going to be as good as what you're inputting," She cautioned.
Which positions data infrastructure as the critical foundation.
"If you're not doing rich, up-to-date structured data, then there's no way you're going to be set up for an agentic world,” said Vallelonga.
He was direct about how Google perceives Australia’s current agentic readiness.
“The area which Australia is really crap at is creatives and feeds. We're the worst in the world at it."
A problem, the panel argued, which being compounded by how brands are treating rising bot traffic.
“For the first time ever, bot traffic is outperforming human traffic,” said Stopps.
“Yet, just over 80% of brands are still blocking some form of AI bot traffic in the back end.”
A misguided instinct, Vallelonga said, as scraper bots could be using the content to surface answers on LLMs.
"Zero click still is a trace. It's not zero," he said.
Have something to say on this? Share your views in the comments section below. Or if you have a news story or tip-off, drop us a line at adnews@yaffa.com.au
Sign up to the AdNews newsletter, like us on Facebook or follow us on Twitter for breaking stories and campaigns throughout the day.