Bronwyn van der Merwe. Credit: Accenture Song
A quarter of executives expect AI to be their biggest audience within three years but brands built on buzz have a problem.
Marketing has spent decades perfecting the art of human persuasion, the emotional hook, the memorable tagline, the campaign that cuts through.
A growing body of evidence suggests a new intermediary is reshaping how purchase decisions get made, and it has no interest in any of that.
ANZ lead at Accenture Song, Bronwyn van der Merwe, said AI agents, the assistants, shopping tools and search experiences increasingly helping consumers research, compare and evaluate options, are becoming a significant force between brands and the people they're trying to reach.
A new Accenture report found a quarter of executives expect these systems to be their single biggest audience for content within three years, ahead of search engines.
Van der Merwe said the shift is already underway in Australia, moving faster than most organisations realise.
"Clear use cases are emerging in areas where consumers face complexity, high levels of choice or ongoing decision-making,” Van der Merwe told AdNews.
“Travel is a good example, where AI can help consumers compare flights, accommodation and activities across multiple providers.
“Retail, beauty, consumer electronics and financial services are also seeing growing adoption because AI can reduce friction and help people make decisions with greater confidence."
The implications for marketing are significant, zero click SEO.
Where a consumer once searched for travel insurance and scrolled through results, they increasingly ask an AI assistant, which evaluates providers and delivers a shortlist, before the customer has visited a single website or seen a single ad.
"Historically, marketers focused on influencing people. Increasingly, they will need to influence both the consumer and the systems helping that consumer make decisions,” Van der Merwe said.
That creates a new and uncomfortable question for brands, does an AI agent actually understand what you offer, and does it represent you accurately when a consumer asks?
Van der Merwe described what she calls a machine-readable brand, one that can be understood consistently by both people and AI systems. The gap between those two things is where many brands currently fall short.
She pointed to a premium mattress brand as a before-and-after illustration.
Before optimising for AI agents, its content does what most brand content does well, lifestyle imagery, emotional testimonials, keyword-rich blogs.
An AI agent evaluating a query like "best mattress for a side sleeper with lower back pain" can't do much with any of that.
After, the same brand structures its content around what an agent needs, standardised firmness ratings, material specifications, trial period and returns policy in plain machine-readable language, works-for and doesn't-work-for guidance, and endorsements from sleep or physiotherapy bodies with linkable sources.
"The result is the brand moves from occasionally mentioned to consistently surfaced with supporting rationale, which in an agent-mediated purchase is the difference between being chosen and being invisible,” Van der Merwe said.
The content infrastructure Van der Merwe said brands need runs across four layers.
The first is structured product data with schema mark up, accessible via API or feed. The second is a claim verification architecture where any superlative, fastest, most effective, highest rated, is backed by linkable, machine-readable evidence.
The third is consistency management across owned content, retailer listings, review platforms and press coverage.
While the fourth is what Van der Merwe calls agent-facing content: structured Q&A that anticipates the questions an agent is likely to ask on a consumer's behalf.
Most organisations, Van der Merwe said, already have the underlying information. The problem is where it lives.
"The challenge is that it's often fragmented across systems and teams. For many businesses, the priority isn't building entirely new capabilities. It's making existing information more connected, accessible and usable,” she said.
The emerging discipline being called generative engine optimisation, GEO, is drawing comparisons to SEO, though van der Merwe sees the objective as meaningfully different. Traditional search returns a list of links.
Generative AI interprets information, compares options and delivers a recommendation directly within a conversation.
"The challenge is no longer just being found but being understood and recommended,” Van der Merwe said.
She expects investment to shift toward content operations, data quality, product information management and measurement.
Gartner has predicted earned media budgets will double by 2027 as a result of GEO's influence, a number van der Merwe flagged as significant for communications teams who have long argued for greater investment in that space.
In ANZ, she says clients in telco, financial services, retail and travel are already moving. Amazon, Etsy and eBay are emerging as early leaders globally, benefiting from strong content, well-structured data and broad distribution across multiple online platforms.
For brands in more commoditised categories, the environment is less forgiving.
Van der Merwe says platforms are likely to carry more influence there, where consumers prioritise efficiency and value.
In high-consideration categories, brands have a clearer path to owning the agent relationship, but only if they invest in first-party data infrastructure, domain expertise and end-to-end orchestration.
The brands most at risk, she says, are those where the gap between promise and delivery is largest.
"AI makes it easier to compare options, verify claims and identify alternatives,” Van der Merwe said.
“Brands with a strong experience foundation, clear information and reliable delivery will be better placed to earn recommendations.
“Those where there's a significant gap between promise and delivery, inconsistent information, unclear policies, poor customer experiences, become much more visible in an agentic environment."
The shift does not, in van der Merwe's view, diminish the value of brand building. It changes where that value sits.
AI systems are becoming effective at processing information and reducing friction. What they don't replicate is trust, emotion, identity and shared experience, the reasons consumers choose brands beyond pure functionality.
"As more routine decisions become automated, the moments consumers choose not to delegate become more valuable,” Van der Merwe said.
For the agency model, she frames the moment as reinvention rather than replacement. Execution, optimisation and content production will increasingly be touched by AI.
Strategic thinking, creativity and customer insight become more important, not less, particularly as brands need partners who can connect marketing, technology, commerce and experience in an increasingly AI-mediated world.
Measurement will also need to catch up.
Van der Merwe points to the concept of share of answer, how often a brand appears in AI-generated recommendations, as an emerging metric alongside traditional reach and conversion.
Marketers, she said, will need visibility into how decisions are being shaped before a consumer takes any action.
"Human behaviour remains the ultimate outcome,” Van der Merwe said.
“What changes is that marketers gain a new layer of visibility into how decisions are being shaped before consumers take action."
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.