New Accenture research shows Australian consumers are increasingly willing to let AI agents discover products and make purchases for them, with 23% already using LLMs as their primary discovery channel. The buying decision is moving inside AI.
When an AI agent picks the shortlist, there is no shelf to stand out on and no ad to run. You are either in the answer or you are not.
New Accenture research says Australian shoppers are ready to hand the buying decision to a machine. Not the browsing. The deciding. 23% of Australian consumers now use large language models as their main way to discover products, and 85% say they would like to shop directly inside generative AI tools.
It goes further than discovery. 62% would rely on AI for at least half of their purchasing decisions. Roughly a third would let an agent make a purchase for them within set limits like budget and brand preference, and around one in ten would let an agent initiate and complete a purchase inside those boundaries.
Then the number that should stop you. 74% said they would trust a personal AI agent more than their best friend to make a purchase on their behalf.
Why it matters
For years the game was getting in front of the customer. The new game is getting in front of the customer's agent, and the agent does not care about your billboard. It cares about what it can find, verify and trust about you across the web.
This is the same story as the AI Overview, one layer deeper. First AI started answering the question instead of sending a click. Now it is being trusted to act on the answer. If the model does not know your brand, cannot verify what you sell, or has learned to describe a competitor more clearly, you never make the shortlist. There is no second page to be found on.
Share of Australian consumers who would trust a personal AI agent more than their best friend to make a purchase for them. Accenture Consumer Pulse research.
What to do about it
Check what the agents already know. Ask the major AI tools to recommend and buy in your category. See whether you appear, and how you are described.
Make your product data machine-readable. Clear specifications, prices, availability and structured data are what an agent reads. Marketing fluff is invisible to it.
Earn third-party corroboration. Agents cross-check. Reviews, credible coverage and consistent listings across the web are what make a model confident enough to recommend you.
Fix your fundamentals before you chase the trend. An agent that reaches a slow, broken or unclear site abandons the task, the same way a person does. The basics you keep putting off are now a discovery problem.
Decide what you will measure. Treat AI visibility and accuracy as a metric with a baseline, not a thing you check once and forget.
The shopfront is not just your website any more. It is how machines describe you when a customer never sees you at all. The businesses that get picked will be the ones that made themselves easy to understand long before the agent came knocking.