Auction-Based Pricing in Retail Media

Auction-based pricing is a pricing model in which competing bids set the price of each ad placement. In retail media it prices sponsored-product clicks on Amazon, Walmart Connect and Instacart: advertisers enter a maximum cost-per-click bid, and the auction decides who wins the placement and what they actually pay.

How auction-based pricing works in retail media

In a retail media ad auction, the advertiser starts by setting a maximum cost-per-click (CPC) bid for an ad group or keyword. That bid, after any adjustments the advertiser switches on, is the ceiling. Amazon states that the final cost per click never exceeds the maximum adjusted bid, and its dynamic bids and placement adjustments can raise the bid itself. The automated vs manual bidding entry covers those adjustments.

When a shopper searches, the retailer's ad server picks the eligible ads and ranks them:

  • Amazon says it cuts the candidates to fewer than 1,000 relevant ads, then scores each one on relevance and bid.
  • Instacart uses CPC bids, product conversion rates and other factors to decide which ads win.
  • Walmart's auction pairs a product's relevancy with its bid. Relevance reflects the match to the shopper's intent, contextual relevancy, item page quality and the product's historical performance at Walmart.

A more relevant ad can therefore win with a lower bid. The winner can also pay less than its bid. Instacart charges $0.01 more than the next-best competitor's bid, and Walmart's winner pays the amount needed to beat the competing relevant bids. Networks also set minimum prices, called reserves or floors, which the floor price entry covers.

Second-price, first-price and fixed prices in retail media

Instacart and Walmart Connect describe their sponsored-product auctions as second-price. Walmart Connect launched an advanced second-price auction for Sponsored Products in 2022, and Instacart states that it uses a second-price model. Amazon says it has priced clicks with a form of generalized second-price auction since it first introduced ads in 2006. The FTC disputes that description, as the example below explains.

Offsite programmatic inventory, including connected TV (CTV), is sold through exchanges that can run first-price auctions, where the winner pays its own bid. The first-price auction entry explains that rule. In OpenRTB, the protocol behind real-time bidding (RTB), the bid request's auction type field names the rule: 1 for first price, 2 for second price plus.

Some inventory skips the open auction. A deal can set its auction type to 3, meaning the floor value is the agreed deal price, and can be flagged as guaranteed. That is fixed price buying inside the same programmatic pipes.

Example: how Amazon says it prices a Sponsored Products click

On August 31, 2026, the FTC and 22 states sued Amazon. They allege that Amazon described its Sponsored Products auction as generalized second-price while, beginning in 2019, a surcharge Amazon referred to internally as a soft reserve price made advertisers pay more than the second-price result. Amazon disputes the claims and published its own account of the auction the same day.

In Amazon's account, its early auctions charged just enough to beat the next highest ranked ad. Each eligible ad now receives a ranking score based on relevance and bid, with relevance weighted more heavily over time. Amazon describes two reserve prices. The hard reserve is the minimum a bid must pass to enter the auction, and Amazon calls the soft reserve a real-time minimum meant to reflect what the placement is worth. Amazon says a winning bid that clears both reserves pays the soft reserve, a winning bid that clears only the hard reserve pays its own amount, and no advertiser pays more than its bid.

Why it matters for brands and sellers

Your maximum bid, after any adjustments, is the ceiling, and competitors and reserves set the price beneath it. Under a second-price rule the runner-up's bid decides what the winner pays, so the same bid costs more on a keyword where many advertisers compete than on one where few do.

Relevance also works at the level of the individual product. Instacart notes that ads in the same ad group still rank differently because each has its own metadata and conversion rates. One keyword bid can buy different positions for different SKUs, and item page quality and historical performance feed Walmart's relevance. Auction rules differ by network, so compare them on the retail media networks hub and size bids with an Amazon keyword research tool. For Walmart specifics, see the Walmart Connect guide.

Frequently asked questions

What is auction-based pricing?

Auction-based pricing is a pricing model in which competing bids set the price of an ad placement. In retail media, advertisers enter a maximum cost-per-click bid for sponsored products, and the network's auction decides who wins and what they pay, which can be less than the bid.

How do retail media ad auctions set the price you pay?

The network ranks eligible ads on bid and relevance and picks a winner. Second-price networks such as Instacart then charge the winner $0.01 more than the next-best bid. Amazon says it charges either a real-time soft reserve or the advertiser's own bid, depending on which reserves the bid clears, a practice the FTC is challenging.

Does the highest bid always win a sponsored product auction?

The winning ad is the one with the best combination of bid and relevance. Amazon says it scores each eligible ad on both, and Walmart pairs a product's relevancy with its bid. Amazon also says highly relevant ads have increasingly won placements with lower bids, so a well-matched product can beat a higher bid.

Is auction-based pricing the same as dynamic pricing?

The two terms describe different prices. Auction-based pricing sets what an advertiser pays for an ad placement, through competing bids in the network's auction. Dynamic pricing concerns the price a shopper pays for the product itself. Retail media teams deal with the first when they bid on sponsored products.

RMIQ enables retailers and brands to master auction-based pricing across retail media by unifying planning, execution, and optimization within a single AI-driven platform that spans Walmart, Instacart, Amazon, Sprouts, Thrive Market, Target, Uber, and more than twenty networks covering up to 85% of the U.S. retail audience. Instead of juggling multiple dashboards and fragmented datasets, your team gains one command center where autonomous multi-agent intelligence continuously adjusts bids, reallocates budgets, orchestrates A/B tests, refines keyword strategies, and propagates cross-network learnings in real time to maximize ROAS and revenue efficiency. Its agents operate at SKU level, aligning item economics, inventory signals, and competitive dynamics with marketplace clearing prices, so your auction posture remains adaptive as demand, seasonality, and rival activity shift by the hour.

With real-time bidding and granular insights, RMIQ calibrates floor and ceiling bids, protects margins, and identifies high-intent queries, while pacing algorithms smooth spend volatility and prevent overpaying during thin liquidity or surge windows. The result is measurable lift—customers report an average 50%+ ROAS increase and up to five dollars in incremental sales per dollar invested—delivered without constant manual oversight. The unified interface consolidates performance dashboards, reporting, and workflows, accelerating decision cycles for both small portfolios and enterprises managing thousands of SKUs, and onboarding is streamlined, with set-up in minutes and responsive customer support.

Governance features enable controlled experimentation and budget stewardship, while explainability surfaces the rationale behind agent actions, helping finance and merchandising stakeholders align on trade-offs between reach, profitability, and share growth. By replacing static rules with continuously learning agents, RMIQ transforms auction-based pricing from a reactive, siloed effort into a proactive, scalable capability that compounds advantages across channels, categories, and campaigns. This positions teams to forecast outcomes more reliably, defend spend in executive reviews, and capitalize on opportunities faster than competitors constrained by manual tooling.

Our Current Partners

We are already helping leading retailers and platforms grow their retail media businesses, including:

Drop your email 

and we’ll show you how to double your retail media ROAS – no strings attached