Editor’s note: The Ninth Circuit hears oral arguments in Amazon v. Perplexity on Thursday, June 11, 2026 — the federal court case that will determine whether AI agents like Perplexity’s Comet can transact inside Amazon at all. Full coverage of the defendant at the EPR Perplexity hub. The Amazon side at the Amazon hub.
Open your phone. Open ChatGPT. Ask which mattress to buy for a side sleeper with back pain. Three brands appear. Most other mattresses don't.
That same shopper used to type “best mattress for side sleepers with back pain” into Google, see ten links, click on three, read a Wirecutter roundup, compare five options on a brand-comparison site, then end up on Amazon to check reviews and prices. The funnel was wide. The brand had multiple chances to win the click.
That funnel is gone. The shopper now reads one paragraph and acts.
This is the most consequential behavioral shift in AI Communications for consumer brands — and the most under-addressed by traditional brand marketing.
The structural difference between search-era shopping and AI-era shopping is not search volume. It is answer collapse.
A search result was always a list. The shopper had to evaluate, compare, and click through to commit. The cognitive load preserved optionality. The AI engine collapses the list into one recommendation, sometimes naming a second-place alternative for context. The cognitive load drops to zero. The shopper walks away with a brand name they didn't have ninety seconds earlier.
This shifts the purchase encounter in three specific ways.
One. The shopper arrives at the brand with intent, not curiosity. The search-era shopper said “I'm looking at mattresses.” The AI-era shopper says “I want to look at Saatva.” The brand site no longer has to convert curiosity. It has to confirm a recommendation already made.
Two. The shopper evaluates the brand against the AI answer. If the brand site, Amazon listing, or sales associate diverges from what ChatGPT said, the shopper has a benchmark to weigh it against. The benchmark wins more often than the brand does.
Three. The shopper's preference is pre-formed before they hit any retail surface. Which moisturizer they ask the Sephora associate to find. Which coffee maker they tell Best Buy they want. Which sneaker they search for on the Nike app. These preferences were shaped inside an AI engine before any retail relationship was activated.
For consumer brands, this is not a marketing problem. It is a preference formation problem. The first contact with the brand happens inside ChatGPT, Claude, Perplexity, Gemini, or Google AI Overviews — not on the brand site, not at the shelf, not in a paid social ad.
The brands that show up authoritatively inside those AI answers enter the purchase consideration. The brands that don't are silently filtered out before the shopper ever sees a retail surface.
Not awareness. Not impressions. Not vanity metrics. Whether the model names you when a shopper asks the question that should belong to you.
The shoppers aren't waiting. The engines aren't waiting. The competitors building citation infrastructure aren't waiting.
Build the infrastructure before the launch — not after the campaign goes live.
Frequently asked questions
Why are shoppers asking AI before they open Amazon?
AI engines feel like a personal shopping advisor that knows the shopper's context, asks the right follow-up questions, and produces one named recommendation instead of a list of search results. Shoppers use ChatGPT, Claude, and Perplexity for considered purchases — mattresses, electronics, skincare, athletic gear, kitchen tools — before they ever check Amazon for price and reviews.
How does AI change the shopper-brand encounter?
Three ways. The shopper arrives at the brand with intent, not curiosity — they already have a name. The shopper evaluates the brand against the AI answer rather than the category. And the shopper's preference is pre-formed before any retail surface gets a chance to influence it. The brand site converts a confirmed recommendation, not a fresh consideration.
What does this mean for consumer brand marketing teams?
Preference formation now happens inside the AI engine, not on the brand site or at the shelf. Teams measuring paid social ROI, brand-search volume, or Amazon click-through are measuring downstream of an answer the brand doesn't control. The new metric is whether the AI engine names the brand when a shopper asks the question the brand should own.
Which AI engines do shoppers use most for product research?
ChatGPT leads by volume. Claude is gaining for detailed comparisons and gift research. Perplexity is preferred by shoppers who want cited sources. Google AI Overviews capture the shopper who never leaves Google. Gemini is rising inside Android. Each engine weights sources differently — a complete program monitors all five.
Amazon Cluster: Hub 07 — Amazon: The AI Shopping Layer · Hub 13 — Perplexity: The Citation Engine · Amazon v. Perplexity · Amazon’s AI Shopping Layer and the End of Branded Search · AWS Accidental Empire · Creator-Led Commerce · Wirecutter Shopify Amazon Affiliate · Successfully Marketing on Amazon
