LEOPRD Investigates AI’s Impact on Brand Visibility in Australian Online Retail
LEOPRD has released a comprehensive study examining the frequency with which brands are omitted or disregarded in AI-driven shopping recommendations, raising concerns about its implications for online retail demand in Australia.
According to the findings, in response to inquiries about unbranded categories, the monitored brand was entirely missing from 36% of the responses, while it was acknowledged but not selected in an additional 35%.
Only 29% of those answers identified the monitored brand as the primary suggestion. This analysis was based on 31,200 AI-generated responses from an expansive monitoring initiative encompassing 27 brands, seven categories, and eight AI platforms.
In parallel with this brand-monitoring venture, LEOPRD scrutinized 420,903 anonymized prompts sourced from Prompt Cowboy, revealing that 44% involved users seeking AI assistance for information gathering, research, advice, or decision-making support.
Additionally, 23% explicitly solicited AI to compare, assess, validate, or recommend a course of action.
Interestingly, Australian users exhibited a higher propensity than their counterparts in the UK and US to consult AI for decision verification.
The research indicated that 16% of Australian prompts requested such validation, in contrast to 13% in the UK and merely 11% in the US.
LEOPRD has formulated a commercial model based on these findings, positing that brand exclusion from AI responses correlates with potential exposure of A$2.9 million for every A$100 million in Australian online retail expenditure, with a sensitivity spectrum ranging from A$2.2 million to A$8.0 million. These figures reflect exposure estimates rather than calculated revenue losses.
Moreover, it was anticipated that approximately A$2.4 billion in annual Australian online retail spending could be exposed within the intersection of AI usage, adherence to recommendations, and brand exclusion.
Decision Support
The report contends that AI is evolving from merely retrieving information to actively influencing consumer choices before shoppers arrive at a retailer’s website or sales channel.
Requests for AI-assisted actions related to vendors, services, or commercial interactions surged dramatically between March and August 2026, escalating from around six to twelve requests per 1,000 prompts.
Such prompts encompassed various tasks, including preparing outreach communications, establishing vendor requirements, structuring interactions with providers, and advancing through purchasing workflows.
Although this behavior remains relatively nascent, the report suggests that it illustrates AI’s progression toward transactional engagement.
When AI provides an erroneous response concerning a brand, there are repercussions. The consumer may squander both time and funds.
Businesses face potential returns, complaints, or support inquiries, and when physical products are ordered, shipped, and returned unnecessarily, there are environmental costs associated with excessive packaging, transportation, and waste, remarked Celia Harding, Founder of LEOPRD.
The report emphasizes that the challenge extends beyond mere factual inaccuracies. It posits that AI systems may also inadvertently disregard a business if a user provides insufficient information or if the public evidence accessible to the AI is too sparse or dubious to endorse a recommendation.
Consumer Trust
LEOPRD also referenced distinct experimental research highlighting how readily individuals yield to AI guidance.
The report indicated that participants adhered to AI recommendations 80% of the time when they were inaccurate and 93% of the time when they were correct, although the tasks involved were centered on experimental reasoning rather than shopping conclusions.
Professor Ben Newell, Director of the Institute for Climate Risk & Response and Professor of Behavioral Science at UNSW Sydney, commented on this tendency.
Individuals are increasingly turning to AI for assistance in framing their needs and evaluating their options.
The risk lies in the fact that a self-assured answer may project an aura of authority, leading people to accept it without scrutinizing the underlying assumptions or evidence. This crosses beyond mere cognitive offloading; it borders on cognitive surrender, noted Newell.
Harding emphasized the significance of this shift, asserting that AI does not need to finalize the purchasing decision to influence consumer considerations.
“AI’s influence doesn’t hinge on making the ultimate decision. If it narrows down the options to three viable companies, reinforces a preference that the consumer already harbors, or offers rationale to exclude another choice, it has effectively affected the commercial outcome,” Harding added.
Wider Responsibility
The report contends that the responsibility for a brand’s visibility in AI-generated responses should not fall solely on marketing teams.
It asserts that the evidence utilized by AI models is derived from various organizational facets, including product data, customer complaint handling, pricing strategies, media coverage, reviews, and commentary from executives.
This implies that enhancing a brand’s representation in AI-driven recommendations may necessitate operational adjustments in addition to changes in content or search strategies.
While marketing teams can optimize content, they cannot rectify systemic product issues, recurrent customer grievances, or deficits in independent evidence.
Organizations must begin to conceptualize AI-generated feedback as a commercial outcome: identifying where they are overlooked, understanding why alternative options prevail, addressing the core evidence, and linking modifications back to demand, returns, complaints, service inquiries, and consumer churn, asserted Harding.
Source link: Ecommercenews.com.au.
