When AI becomes the intermediary between a brand and its buyer, the recommendation is not shaped by who spent the most on placement. It is shaped by what people actually said in reviews, what independent tests found and what earned coverage reported.
That is the argument made by Inside Retail Australia this week, and the data backs it up. ChatGPT references product reviews in 58% of its responses. Perplexity cites review platforms in 100% of product-related answers. The entire architecture of modern brand marketing, built on the assumption that attention can be purchased and perception can be manufactured, now runs into a system that simply does not care about your ad spend.
Of ChatGPT responses reference product reviews when answering purchase-related queries
The mechanism
AI systems pull from review platforms like G2, Capterra, TrustRadius, Google Reviews and Amazon ratings. They weight these sources more heavily than brand-owned content because they represent independent validation. A brand page says "we are the best." A thousand four-star reviews say "they are consistently good." AI trusts the second signal.
This means a business with average marketing but excellent reviews will increasingly outrank a business with exceptional marketing and mediocre reviews. The hierarchy has inverted.
The trust problem
Nearly six in ten shoppers (58%) blame the retailer or brand when an AI recommendation contains incorrect product information. Sixteen percent say they would avoid purchasing the product entirely after a bad AI recommendation. The margin for error is thin and the consequences are immediate.
For brands that have relied on paid placement, influencer partnerships and creative campaigns to compensate for an average product experience, this is a structural problem. You cannot buy your way past a three-star aggregate rating when the AI is reading the reviews before the customer does.
Why it matters
Product quality has always mattered. What has changed is the mechanism of discovery. Previously, a customer might see an ad, visit a site, read curated testimonials and convert without ever encountering the aggregate review sentiment. Now, the AI synthesises that sentiment before the customer even reaches your site.
This is especially relevant for Australian ecommerce brands competing in categories with strong international competition. If a competitor has better reviews and the AI is recommending based on review quality, no amount of local brand building compensates.
What to do about it
The brands that win in AI-mediated discovery will be the ones that earned the reviews, not the ones that bought the impressions.
