Sponsored listings are a ripoff…for sellers

Welcome to the age of danegeld.

Cory Doctorow
4 min readNov 29, 2023
A woodcut illustration of Britons paying Danegeld to Viking conquerors. Most of the scene has been faded back and washed out, but the loot, the supplicant and the Viking lord have all be colorized.

Tonight (November 29), I’m at NYC’s Strand Books with my novel The Lost Cause, a solarpunk tale of hope and danger that Rebecca Solnit called “completely delightful.”

Not all ads are created equally sleazy. The privacy harms from surveillance ads, though real, are often hard to pin down. But there’s another kind of ad — or “ad” that picks your pocket every time you use an ecommerce site.

This is the “sponsored listing” ad, which allows merchants to bid to be among the top-ranked items in response to your searches — whether or not their products are a good match for your query. These aren’t “ads” in the way that, say, a Facebook ad is an ad. These are more #payola, a form of bribery that’s actually a crime (but not when Amazon does it):

https://en.wikipedia.org/wiki/Payola#U.S._investigations_and_aftermath

Amazon is the global champion of payola. It boasts of $31 billion in annual “ad” revenue. That’s $31 billion that Amazon sellers have to recoup from you. But Amazon’s use of “most favored nation” deals (which requires sellers to offer their lowest prices on Amazon) mean that you don’t see those price-hikes because sellers raise their prices everywhere:

https://pluralistic.net/2023/04/25/greedflation/#commissar-bezos

Forget Twitter: Amazon search is the poster-child for enshittification, in which Amazon locks you in (for example, with a year’s shipping prepaid through Prime) and then you get recommended worse products while sellers make less money and Amazon pockets the difference.

Sellers who don’t sell on Amazon are dead in the water, because most US households have Amazon Prime and overwhelmingly, Prime users start their search on Amazon, and, if they find the goods they’re seeking. After all, they’ve prepaid for shipping.

So sellers suck it up and pay a 45–51% Amazon tax and pass it on to us — no matter where we shop. A lot of the junk fees sellers pay are related to Prime and other fulfillment services, but an increasing share of the Amazon tax comes from the need to pay to “advertise,” because if they don’t buy the top result for searches for their own products, their competitors’ ads will push them right off the first page (those competitors spend money on advertising, rather than manufacturing quality).

There’s a lot of YOLO/ROFLMAO in those ads: search for “cat beds” and 50% of the first five screens are ads — including ads for dog products, apparently bought by companies adopting a spray-and-pray approach to advertising. Someone selling a quality product still has to outbid all of those garbage sellers:

https://pluralistic.net/2022/11/28/enshittification/#relentless-payola

This is at the root of Amazon’s Pricing Paradox: while Amazon can defend itself against regulators by citing sellers whose prices are lower and/or whose quality is higher, it’s nearly impossible for shoppers to get those deals. If you click the top result for your search, you will, on average, pay 29% more than you would if you found the best bargain on the site:

https://pluralistic.net/2023/11/06/attention-rents/#consumer-welfare-queens

What’s more, you can’t fix this by simply sorting by price, or by reviews, or some mix of the two. The sleaziest sellers have mastered tricks like changing the number of units they sell so the total price is lower. For example, if batteries are normally sold $10 for a four-pack, a sleazy seller can offer batteries at $9 for three units. A lowest-to-highest price-sort will put this item ahead of a cheaper rival.

Researchers found that getting a good deal at Amazon requires that you make a multifactorial spreadsheet by laboriously copy/pasting multiple details from individual listing pages and then doing sorts that Amazon itself doesn’t permit:

https://scholarship.law.bu.edu/faculty_scholarship/3645/

There’s an exception to this: Amazon and Apple have a cozy, secret arrangement to exclude these “ads” from searches for Apple products. But if you’re shopping for anything else, you’re SOL:

https://www.businessinsider.com/amazon-gives-apple-special-treatment-while-others-suffer-junk-ads-2023-11

These payola markets are bad for buyers, and they cost sellers a lot of money, but are they at least good for sellers? A new study from three business-school researchers — Vibhanshu Abhishek, Jiaqi Shi and Mingyu Joo — shows that payola is a very bad deal for good sellers, too:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3896716

After doing a lot of impressive quantitative work, the authors conclude that for good sellers, showing up as a sponsored listing makes buyers trust their products less than if they floated to the top of the results “organically.” This means that buying an ad makes your product less attractive than not buying an ad.

The exception is sellers who have bad products — products that wouldn’t rise to the top of the results on their own merits. The study finds that if you buy your mediocre product’s way to the top of the results, buyers trust it more than they would if they found it buried deep on page eleventy-million, to which its poor reviews, quality or price would normally banish it.

But of course, if you’re one of those good sellers, you can’t simply opt not to buy an ad, even though seeing it with the little “AD” marker in the thumbnail makes your product less attractive to shoppers. If you don’t pay the danegeld, your product will be pushed down by the inferior products whose sellers are only too happy to pay ransom.

It’s a system where everybody loses — except monopoly ecommerce platforms, who enshittify everything and rake it in.

If you’d like an essay-formatted version of this post to read or share, here’s a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:

https://pluralistic.net/2023/11/29/aethelred-the-unready/#not-one-penny-for-tribute

--

--