The first thing Amazon tells every new employee is also the thing its advertising division has quietly stopped doing.
"Leaders start with the customer and work backwards." That is Amazon's Leadership Principle number one - customer obsession - the north star that Jeff Bezos embedded into the company's DNA before the word "flywheel" became a business school cliche. It is posted on the corporate jobs site. Andy Jassy, Bezos' successor, filmed an hour-long tutorial on it. It is, in theory, the reason Amazon became the most dominant retailer in human history.
Here is what happens when that principle runs headfirst into a $56 billion revenue line: the principle loses.
Amazon Ads - the part of the company that sells advertising space on amazon.com - generated $56.2 billion in revenue in 2024, growing 18% year over year. That makes it the third-largest digital advertising business in the world, behind only Google and Meta. And a significant portion of that money comes from a mechanism that is, depending on your perspective, either brilliant or quietly corrosive: selling the first three spots on any search results page to whoever bids the most, with no obligation to tell the shopper that is what they are looking at.
When a person types "shaver" into Amazon's search bar, the first products that appear are not the best shavers. They are the shavers whose manufacturers paid the most to be there that day. The distinction, for most shoppers, is invisible.
The background
Amazon was founded in 1994 as an online bookstore. Within a decade, it had pivoted to selling everything, and within two decades it had figured out something that most retailers still have not: the most valuable real estate in commerce is not a shelf in a store. It is the top of a search results page.
The flywheel - a term Bezos used to describe Amazon's self-reinforcing business model - worked like this: lower prices attract more customers, more customers attract more sellers, more sellers create more competition, which drives prices lower still. The wheel keeps spinning. Every feature Amazon added was supposed to feed the flywheel. Faster shipping fed it. Better reviews fed it. The Amazon's Choice badge - a small label applied to certain products indicating they were highly rated, frequently purchased, and rarely returned - fed it by helping overwhelmed shoppers make faster decisions.
Badges like "Best Seller" and "Amazon's Choice" are examples of what retail designers call social proof - shorthand signals that save the shopper from having to read every product listing. They work because they are, at least in theory, earned. The shopper sees the badge and trusts that something independent has decided this product is worth buying.
The advertising business grew alongside all of this. At first it was modest - a way for sellers to boost visibility. Then it became enormous. Amazon introduced sponsored products - ads that look almost identical to organic search results but pay for their placement - and the money poured in. By 2024, 99% of Amazon searches showed sponsored products, with an average of 20 sponsored listings appearing per page load. The ad machine had eaten the search results page.
What is actually happening
Jean-Christophe Solus, a shopper experience consultant who has studied Amazon since its founding year of 1995, published a detailed critique of Amazon Ads in May 2025 that identifies three specific ways the advertising division is quietly undermining the shopping experience the company built its reputation on.
The first is the badge problem. Amazon's Choice is algorithmically awarded - it goes to products with high ratings, frequent purchases, and low return rates. That is a reasonable set of criteria. The problem, as Solus points out, is that for years Amazon never explained this to shoppers. The badge just appeared, and shoppers were expected to trust it without knowing what it meant. Amazon has recently begun adding explanations to some badges - a small improvement - but the "Best Seller" and "Exclusive" labels still carry no transparent criteria. A shopper seeing "Best Seller" in a product thumbnail has no way of knowing whether that means best-selling in the past hour, in a broad category, or in a narrowly defined subcategory created specifically to make the label easier to earn.
The second is what Solus calls ghost brand pages. Amazon allows manufacturers to publish rich, brand-designed content on product pages - detailed imagery, storytelling, technical specifications formatted in the brand's visual language. This was partly a strategic response to Shopify, which built its entire business on offering brands the kind of ownership over their presentation that Amazon's generic product pages could never match. Amazon's version is free and often well-produced. The problem is that it is buried. After 30 years of growth, Amazon product pages have become extremely long, with the manufacturer's premium content lost somewhere below the fold amid reviews, Q&As, comparison tables, and more ads. There is no navigation element that lets a shopper actively seek out this content.
The third problem is the most consequential. A large proportion of Amazon's advertising revenue comes from sponsored ads - paid placements that appear at the top of search results. These ads carry the word "Sponsored" in small type, but they make no claim about the product being advertised. The ad does not say "this shaver is the closest shave on the market" or "this battery lasts 15 hours longer." It simply pays to be visible. The product is there because money changed hands, not because Amazon has assessed it as the best answer to the query.
This is what Solus calls claimless advertising - ads that exist purely to capture attention at the point of highest purchase intent, without providing any information that helps the shopper decide. According to eMarketer, over two-thirds of Amazon's US ad revenues come from search ads. The entire architecture of Amazon's most profitable ad format is built on placement without promise.
The money trail
The number that explains everything here is $56 billion. That is what Amazon Ads made in 2024. To put it in context: that is roughly the entire annual GDP of a country like Croatia. It is more than the combined advertising revenue of every newspaper and magazine in the United States. Amazon's ad segment now represents approximately 9% of its total revenueand carries operating margins significantly higher than its core retail business, which operates on razor-thin margins to maintain the low-price positioning that attracts shoppers in the first place.
Here is the structural tension that makes this hard to fix. Amazon's retail business depends on offering the lowest prices and the best selection. Its advertising business depends on charging sellers for visibility regardless of price or quality. These two objectives are not inherently incompatible - but they pull in opposite directions when the question is whether a shopper buying a battery is seeing the best battery or the most aggressively advertised one.
The sellers paying for sponsored ads are not doing so out of altruism. They pay because the cost-per-click model - where advertisers pay only when a shopper clicks their ad - makes the return on investment calculable and often very attractive. A seller paying $0.78 per click who converts one in fifteen shoppers and makes $12 profit per sale is running a positive return. The maths works. But the maths working for the seller does not mean it works for the shopper, who may be clicking a product that ranked third on quality metrics had the results been purely organic.
The flywheel, in other words, has developed a leak. The original model fed on trust - shoppers returned because the results were good, which attracted more sellers, which created more competition, which improved quality. The advertising model feeds on something different: visibility purchased at auction. A well-funded but mediocre product can outrank a better product with less marketing budget, consistently and at scale.
Walmart, the largest physical retailer in the world, has adopted the same approach. As Solus notes, the claimless sponsored ad format has spread across retail search. Once Amazon demonstrated that shoppers would tolerate it - or more precisely, would not notice it - the incentive for competitors to do otherwise collapsed.
What people are doing about it
Some sellers are pushing back through their budgets. Small brands with limited advertising spend are increasingly looking for visibility through organic ranking improvements - investing in reviews, fulfilment speed, and pricing rather than sponsored placement. The calculation is that the auction for sponsored spots in competitive categories has become expensive enough that the return no longer justifies the cost.
Some larger brands are hedging by investing simultaneously in their own direct-to-consumer (D2C) channels - websites they control entirely, where the search results show exactly what the brand wants, and where no competitor can pay to appear above them. Shopify has benefited significantly from this dynamic; brands that once sold exclusively through Amazon are building parallel storefronts that are not subject to Amazon's advertising auction.
Amazon itself has made a small concession in the badge transparency direction - adding explanatory pop-ups to some product labels that describe the criteria behind the designation. It is the kind of reform that is visible enough to cite but narrow enough not to cost the company much. The core sponsored ad model, where the top results are paid placements with no claim requirement, remains unchanged.
Regulatory pressure is building on a related front. The FTC sued Amazon in 2023 over its Prime subscription cancellation process, arguing the company used manipulative practices to trap customers into renewals. The suit tested the credibility of the customer obsession principle in a federal courtroom. The outcome of advertising-specific regulatory scrutiny, particularly around transparency of paid search placement, remains an open question in the US and Europe.
Solus's proposed fix is modest and technically achievable: pair paid advertising placements with a brand-selected badge that states a specific, verifiable claim about the product. The ad slot would still be sold. The revenue would still flow. But the shopper would receive something in return - an actual reason to click, rather than just a position purchased at auction.
The bottom line
Amazon built its reputation on the idea that it is always on the shopper's side. That principle is still posted on the wall. But Amazon Ads, now a $56 billion business growing at 18% per year, operates on a fundamentally different logic: the seller who pays the most appears first, with no obligation to explain why the shopper should care. The badges that were designed to create trust have become ambiguous. The premium brand content that was designed to inform shoppers is buried. And the ad format that generates the most revenue tells the shopper nothing except that someone spent money to reach them. The flywheel still spins. It is just spinning for different people now.
Timeline
- 1994 - Jeff Bezos founds Amazon as an online bookstore, beginning the experiment in customer-first retail design
- 1995 - Amazon.com opens to the public; Jean-Christophe Solus begins tracking its retail model
- Early 2000s - Amazon introduces the "Amazon's Choice" and "Best Seller" badges to help shoppers navigate an expanding product catalogue
- 2005 - Amazon Prime launches, beginning the loyalty ecosystem that would eventually become the backbone of the advertising audience
- 2012 - Amazon Advertising (then Amazon Marketing Services) begins expanding sponsored product placements across search results
- 2024, January - Amazon introduces advertising on Prime Video by default, adding streaming inventory to the advertising platform and reaching over 200 million viewers worldwide
- 2024, full year - Amazon Ads generates $56.2 billion in revenue, an 18% increase year over year, making it the third-largest digital ad business globally
- 2024, Q4 - Amazon's ad revenue reaches $17.29 billion for the quarter, up 18% year over year; CEO Andy Jassy cites a $69 billion annual run rate
- 2025, Q1 - Amazon Ads generates $13.9 billion, up 18% year over year, outpacing overall company revenue growth of 9%
- May 2025 - Shopper experience consultant Jean-Christophe Solus publishes a critique of Amazon Ads, arguing the business violates Amazon's own customer obsession principle through claimless ads, opaque badges, and buried brand content
Summary
Who: Amazon Ads, the advertising division of Amazon.com, and the brands and sellers who pay for placement on its platform
What: A $56 billion advertising business built on sponsored search placements that prioritise paid visibility over product quality, undermining the trust signals - badges, brand pages, and search results - that Amazon's retail reputation was built on
When: The tension has been building since Amazon began scaling its advertising business in the early 2010s; the critique published in May 2025 crystallises the accumulated drift from the original customer-first model
Where: Primarily amazon.com search results pages, where sponsored products now appear in 99% of searches with an average of 20 paid placements per page load
Why: Amazon's advertising business operates on fundamentally different incentives from its retail business - sellers pay for placement regardless of product quality, creating a structural conflict with the customer obsession principle that Amazon publicly holds as its founding value