Can Google's pricing tool really prevent price increases?

Matt Stoller's piece also argues that Google's reassurance about its new pricing tool is misleading. After pricing expert Lindsay Owens — who "helped uncover the dynamic pricing scheme of Instacart" — warned the tool could let retailers "overcharge" consumers using their own data, Google responded that the service lets merchants "offer a lower priced deal or add extra services like free shipping — it cannot be used to raise prices." Stoller rejects that framing: "the idea that discounts mean lower price levels is foolish." He points to Google's own statement that the feature is meant "to increase revenue generated from consumers, to 'help shoppers prioritize value over price alone.'"

His reasoning is an analogy to existing discount-washing practices. A hospital bill, he notes, might list an MRI at $8,000, apply a $6,000 "discount," and settle at a cash price of $1,970, of which the patient's billed portion is $30 — "that $8000 number is fake, so is the discount, the only thing that matters is the cash changing hands." He extends the same logic to pharmacy benefit managers, which "get 'discounts' off of ludicrously high list prices," and to department stores that permanently mark goods "80% off." Applied to Google's tool, the argument is that a "discount" is only meaningful relative to a real reference price; if merchants can set the list price arbitrarily high before Google applies a personalized "deal," the promise that the system "cannot be used to raise prices" describes the sticker-price mechanics, not what the consumer actually pays.

This complements Can an AI agent serve both merchant and user interests fairly?: both describe a commercial intermediary whose monetization model — referral fees for Azhar's agent, merchant payments for "Direct Offers" here — creates an incentive to extract value from the buyer side even while appearing to help them. It also stands in tension with Does higher purchase volume make AI shoppers search less carefully?: that study found a Gemini-based agent degrade toward simple price-minimization under certain conditions, whereas Stoller's concern is the mirror failure — a Gemini-powered system marketed as helping shoppers find "value," which Google itself says is meant to raise what it extracts from them.

The excerpt offers no data on how Direct Offers' discounts compare to baseline prices in practice, and Stoller cites no study showing consumers paying more; the claim rests on analogy to known discount-washing patterns in healthcare and pharmacy pricing, made in an opinion newsletter rather than a regulatory finding. The implication, at the strength the analogy supports, is that a cannot raise prices promise is unfalsifiable without visibility into how the reference list price itself gets set, so reviewers of agentic pricing tools should treat list-price inflation, not just the nominal discount, as the thing to audit.