Ad Tech Briefing: Amazon’s auction trial puts agentic advertising’s trust problem in focus

According to the FTC, Amazon imposed surcharges of up to 50% on ordinary shopping days and 250% during major retail events, generating an additional $87 million during the 2022 holiday season. Amazon disputes the allegations, and the legal proceedings are ongoing.

Amazon recently reported quarterly advertising revenue approaching $20 billion, with  Amazon Ads Agent – an AI-powered platform that’s likely to attract accusations of “black box” from detractors – a key part of its push to automate campaign planning, buying and optimization.

And with Amazon, the third pillar of adland’s triopoly, the key question for advertisers is whether they should hand over more control to a platform when they cannot independently verify how its existing auctions determine prices.

FTC allegations

The FTC and 22 state attorneys general accuse Amazon of misleading advertisers about its Sponsored Products auctions.

Amazon presented these as generalized second-price auctions, where advertisers generally pay prices determined by competing bids rather than their maximum bid.

However, regulators allege Amazon introduced undisclosed reserve-price mechanisms that increased charges without corresponding competition, effectively pushing prices closer to first-price auctions.

According to the government body’s allegations, during the 2021 holiday season, Amazon increased average cost per click by 20%, with some advertisers seeing increases exceeding 200%. The complaint also alleges executives considered using reserve prices to address a $40 million revenue shortfall.

The allegations extend to how Amazon explained rising prices, i.e., whether Amazon misrepresented how its auctions worked and what advertisers were charged.

However, Amazon maintains that reserve prices reflect the value of its advertising placements, disputes the FTC’s interpretation, and argues that reserve pricing is not inherently unlawful.

For advertisers, the concern is straightforward: they can see what they paid, but cannot necessarily establish how that price was calculated.

Amazon’s AI ambitions 

The allegations come shortly after Amazon introduced new products designed to automate more advertising decisions.

At October’s unBoxed conference, Amazon unveiled Ads Agent and Full-Funnel Campaigns, allowing advertisers to set objectives, products, creative and budgets while its technology handles targeting, allocation and optimization.

Amazon says beta testing delivered up to 67% higher return on long-term ad spend and 29% lower costs per new-to-brand customer compared with manually configured campaigns. Those figures are company-reported.

The proposition is that advertisers can reduce manual campaign management and improve performance. However, greater automation also means relying more heavily on Amazon to explain how budgets are spent and why its systems make particular decisions.

While the FTC’s allegations concern Sponsored Products auctions, not Amazon’s new AI products, they still add weight to concerns raised about advertisers’ ability to inspect automated decisions, verify results, and understand fees.

The Google comparison

Google’s ad tech antitrust case provides a precedent, although the conduct and markets differ.

A federal court found Google illegally monopolized publisher ad-serving and ad exchange markets – in a separate case, its search practices were also found to be monopolistic. The subsequent remedies stopped short of divestiture but imposed measures intended to improve competition and interoperability.

Both cases concern platforms whose control over advertising infrastructure limits customers’ ability to scrutinize how transactions work.

The difference is that Amazon is now asking advertisers to give its technology greater control over buying decisions, even as regulators challenge the transparency of its existing auction practices. These developments involve different businesses and issues, but they point to a broader accountability problem in digital advertising.

What we heard

“We’re having to look much more closely at what happens after the click and whether those clicks are actually producing qualified leads and revenue.”

— Lindsay Marty, founder of Above the Bar Marketing, speaking to Digiday about advertisers reassessing their Google Search budgets.

Numbers to know

ChatGPT’s ad offering may be on the rise, but the nature of its value is still hard to determine, according to numbers analyzed in recent studies. 

• 6,333: Distinct brands Adthena tracked advertising on ChatGPT in August, up from 1,082 in June, across the U.S., U.K., EU and Australia.
• 43.4%: Proportion of ad-eligible ChatGPT users in the U.S. reached by advertising, according to Similarweb and Graphite, compared with 45.7% in the U.K. and 27.8% in Mexico.
• $0.89: Revenue generated per paid ChatGPT visit for U.S. ecommerce advertisers, compared with $2.45 from paid Google Search, according to research by Maximilian Kaiser and Christian Schulze.
• 0.37%: Conversion rate for paid ChatGPT traffic, compared with 1.17% for paid Google Search, according to the same working paper, which remains under review.

What we’ve covered

⁠Amazon’s Ads nears $20 billion in Q3 as it bets on agentic AI to fuel further growth

Amazon’s new automated buying tools promise simplified execution while raising questions about transparency and advertiser control.

⁠Ad Tech Briefing: Google’s AI-reckoning

Alphabet’s latest earnings reveal how AI is complicating the economics of search advertising even as its cloud business expands.

What we’re reading

⁠Taboola faces Yahoo mediation over its 30-year native ad deal 

A dispute involving one of Taboola’s most important commercial relationships emerges alongside disappointing quarterly revenue.

AI-driven tech layoffs suck, but might lead to certification 

Former Google ads leader, Jonathan Bellack, on his Platformocracy outlet, examines how AI coding threatens junior engineering career paths and argues that professional certification could become increasingly important.

⁠Agencies find AI overuse is leading to brief bloat 

Agency executives warn that AI-generated briefs are creating more work rather than improving productivity.

Media Briefing: Publishers’ revenue growth defies traffic declines, but tougher comparisons loom 

Major publishers continue to grow revenue despite deteriorating referral traffic and mounting AI disruption.