Retail Media Breakfast Club

Amazon already captures roughly 40% of every dollar spent online in the US, yet it continues to grow at an impressive pace. So I wanted to look at a question we don't hear nearly enough: where does Amazon spend its own advertising money? Because while Amazon gives brands advertising recommendations every day, I think its own media plan may be the more interesting lesson.
In this recap of my recent article for The Drum, I dug into three years of Sensor Tower data to see where Amazon is putting its advertising dollars, which channels are getting the biggest investment, and how its mix is changing. I also look at the shift from desktop display to video, why Amazon is willing to pay more for potentially fewer impressions, and what advertisers can learn from the company's approach to emerging channels.

This episode is sponsored by GrowthLoop
Timeline
[02:09] - Breaking down Amazon's $2.7 billion US advertising spend in 2025
[03:54] - Why Amazon's media plan looks surprisingly conventional
[07:23] - How Amazon can measure Meta impressions against actual Amazon purchases
[08:00] - What Amazon is actually advertising, from AWS and Prime to devices and the retail store
[09:10] - The striking gap between Amazon's rising ad spend and relatively flat impression growth
[10:20] - Why Amazon is shifting dramatically from desktop display into video
[12:15] - The three habits I see in Amazon's media buying strategy

Links & Resources
Read my full article on The Drum: The best advertising advice Amazon gives isn't in a slide deck
Check out Sensor Tower, who measure the full digital consumer journey, from mobile apps to web, ads, gaming, and beyond.
Follow Ross Walker, Head of Retail Media @ Acadia, on LinkedIn
Read my related articles:Have Retailers Lost Discovery to AI for Good?
Some Retailers Want to Power Other Retailers' Ad Businesses. Will It work?
Amazon Q2 Earnings: What Industry Experts Say About Power Buying, Ad Overload, and AI Threats

Check out Basket Case, a project I’m part of with The Drum and QSIC, an irreverent and lively podcast series talking shop: latest episode
I'm sharing comedy skits on Instagram about retail media! Follow along: @retailmediabreakfastclub
Subscribe to Retail Media Breakfast Club's daily newsletter
Follow Kiri Masters on LinkedIn

Advertising has always been the glamorous side of ecommerce, while product catalog maintenance, reviews, inventory, and other operational work have often been underfunded. But with Amazon’s Sponsored Prompts, that relationship is changing dramatically. This is Part 2 of my Sponsored Prompts series (check out Park 1 here), I’m digging into why your advertising account is now downstream of everything happening in your operations, and why content is no longer just an input to your ads; it is the ad. 
I’m also taking a closer look at product reviews, reporting limitations, bidding, and what Sponsored Prompts could ultimately mean for retail media teams. There are some legitimate concerns here, particularly around review manipulation and transparency, but I also see a fascinating upside: Sponsored Prompts could become a surprisingly powerful form of consumer research, as they could reveal the questions, objections, and selling points that actually convert.

This episode is sponsored by GrowthLoop
Timeline
[01:37] - Why your advertising account is now downstream of operations, including inventory, reviews, commercial terms, and product content. 
[03:18] - How Sponsored Prompts turns product content and backend inputs into the ad itself. 
[05:13] - Why fake and manipulated product reviews could become an even bigger problem when reviews are effectively being used as paid ad copy. 
[09:26] - Amazon advertising: bidding still works the same way, but advertisers now have less leverage at the auction level. 
[10:15] - Why Sponsored Prompts reporting is still surprisingly basic, and what Amazon’s slow rollout might tell us about the ad format. 
[12:49] - The unexpected upside: how Sponsored Prompts could uncover consumer questions and selling points that brands didn't realize they were underselling.

Links & Resources
Check out Part 1 of my Amazon Sponsored Prompts series: Billed for the Ad Before the Ad Exists: Amazon Sponsored Prompts
Ladi Fagbola, an ecommerce account director at the agency Monks, articulated a cool follow-up from my post this week
Follow Ladi Fagbola on LinkedIn
Read my related articles:Billed for the Ad Before the Ad Exists: Amazon Sponsored Prompts
What it means for retail media when AI chat has only 2 ad slots
Amazon Unboxed 2025: Key Announcements & Takeaways for Brand Advertisers
Retailer AI Assistants: Are Shoppers Actually Using Them?
Walmart's Sparky AI Assistant: A Late Mover with Early Promise

Check out Basket Case, a project I’m part of with The Drum and QSIC, an irreverent and lively podcast series where we talk shop.
I'm sharing comedy skits on Instagram about retail media! Follow along: @retailmediabreakfastclub
Subscribe to Retail Media Breakfast Club's daily newsletter
Follow Kiri Masters on LinkedIn

There’s a particular kind of vertigo that comes from watching a podcast you started get better without you. I launched the Ecommerce Braintrust podcast back in 2017, and while I’m no longer running it, I still listen because the team at Acadia is in the trenches every day doing the work.
A recent conversation between my old Acadia colleagues on measurement really got me thinking about a question I hear from brands all the time: if every advertising platform says it’s winning, why don’t the business results reflect that? When Meta, Google, Amazon and retail media networks all claim credit for conversions, the numbers can quickly become impossible to reconcile. In this episode, I share snippets from a recent Ecommerce Braintrust episode that dig into the idea of “engine inflation,” why multi-touch attribution is struggling in 2026, and what brands should be using instead to understand what is actually driving incremental growth.

This episode is sponsored by GrowthLoop
Timeline
[00:00] - The measurement problem every advertiser is wrestling with: when every platform claims to be winning, why don't the business results add up?
[01:45] - Introducing “engine inflation” and why platform-reported revenue can exceed actual business revenue.
[03:56] - Why multi-touch attribution isn't coming to save us and why Acadia has moved away from MTA.
[04:52] - The three-pillar measurement framework: media mix modeling, incrementality testing, and platform optimization.
[06:15] - Why the growth of retail media networks makes a holistic view of total media spend increasingly important.
[08:34] - Why measuring Amazon and Walmart simply by retail media network and ad type can produce the wrong conclusions.
[09:46] - How defensive branded search can look incredibly efficient while doing very little to drive incremental growth.

Links & Resources
Listen to the full Ecommerce Braintrust episode: The 2026 Measurement Crisis: Why Your Media Efficiency is Probably a Lie With Sally Kazin and Ross Walker - Episode 438
Subscribe to Ecommerce Braintrust on Apple Podcasts
Follow Sally Kazin, Head of Analytics @ Acadia, on LinkedIn
Follow Ross Walker, Head of Retail Media @ Acadia, on LinkedIn
Read my related articles:Retail Media's Measurement Problem: It's Not Just the Retailers
'Not 100% altruistic': ad measurement veteran on what retail media is really optimizing for
As Retail Media Spending Soars, Brands Face A Measurement Crisis
The Other Side of the Story: Why Retailers Struggle with Media Measurement & Standardization

Check out Basket Case, a project I’m part of with The Drum and QSIC, an irreverent and lively podcast series where we talk shop.
I'm sharing comedy skits on Instagram about retail media! Follow along: @retailmediabreakfastclub
Subscribe to Retail Media Breakfast Club's daily newsletter
Follow Kiri Masters on LinkedIn

When Amazon unveiled Sponsored Prompts at Unboxed, I initially dismissed them as a clever but relatively minor addition to Alexa Shopping. But after digging into Andrew Bell's deep analysis of the underlying patent, I realized this format represents something much bigger and potentially far more disruptive.
In this episode, I unpack why Amazon's new AI-powered ad format flips decades of digital advertising on its head by charging advertisers before the persuasive content is even generated! I explore what that means for brands, legal teams, product claims, customer reviews, and the future of retail media. This is Part 1 of a two-part series examining one of the most important shifts in AI advertising that brands should already be paying attention to.

This episode is sponsored by GrowthLoop
Timeline
[00:15] Why Amazon Sponsored Prompts seemed simple and why I completely underestimated their impact.
[01:30] How Sponsored Prompts fundamentally reverse the traditional advertising model.
[03:18] Why Amazon's "charge first, generate later" approach is unlike Google's dynamic advertising.
[05:15] The legal and brand risks of AI-generated answers built from reviews and Amazon's own data.
[06:00] Why advertisers may be shifting from buying attention to influencing AI-generated persuasion.
[08:22] Is Sponsored Prompts the future of AI advertising or just an experiment?

Links & Resources
Last week, Andrew Bell did a massive deep dive into the patent behind Sponsored Prompts which changed my thinking about how they will work (Disclosure: Bell is VP of Research at ReFiBuy, a company I advise).
Read my related articles:Amazon Unboxed 2025: Key Announcements & Takeaways for Brand Advertisers
Retailer AI Assistants: Are Shoppers Actually Using Them?
Walmart's Sparky AI Assistant: A Late Mover with Early Promise

I'm sharing comedy skits on Instagram about retail media! Follow along: @retailmediabreakfastclub
Subscribe to Retail Media Breakfast Club's daily newsletter
Follow Kiri Masters on LinkedIn

What makes one retail media network stand out when so many seem to promise the same thing? In this episode, I explore a fascinating psychology concept called the 'effort heuristic', and explain why it has huge implications for retail media. When buyers can't easily judge technical quality, they look for visible signs that a retailer is investing, innovating, and genuinely building something worthwhile.
I also share why transparency can become a competitive advantage, highlight Costco's retail media strategy as a standout example, and discuss where the line is between showcasing meaningful progress and creating announcement fatigue. If you're building, selling, or buying retail media, this episode offers a fresh perspective on what actually influences decision-making before performance data ever enters the conversation.

This episode is sponsored by GrowthLoop
Timeline
[00:00] The psychology study behind the effort heuristic, and why perceived effort influences perceived quality.
[01:47] How the effort heuristic applies to retail media buyers evaluating competing retail media networks.
[03:19] Why Costco's transparent approach to its retail media tech stack stood out, and what other retailers can learn from it.
[05:43] The danger of over-marketing: when every product update becomes an "industry-first" announcement.
[06:33] Why the largest retail media networks play by different rules than emerging RMNs competing for test budgets.
[07:21] My advice for retailers and technology partners: show your work, earn credibility, and build trust through transparency.

Links & Resources
The Effort Heuristic by Kruger, Wirtz, Van Boven and Altermatt.
Read my related articles:Costco Is The First US RMN To Reveal Its Entire Tech Stack
Why RMNs Keep Their Tech Stacks Secret
The ‘Costco Velocity’ Tech Stack Is Coming Together. Here's How the Pieces Fit.

Check out Basket Case, a project I’m part of with The Drum and QSIC, an irreverent and lively podcast series talking shop: latest episode
I'm sharing comedy skits on Instagram about retail media! Follow along: @retailmediabreakfastclub
Subscribe to Retail Media Breakfast Club's daily newsletter
Follow Kiri Masters on LinkedIn