Originally published at: https://pluralistic.net/2023/09/03/big-tech-cant-stop-telling-on-itself/
In July, the Federal Trade Commission announced a complaint against Amazon over the ways the company has tricked customers into subscribing to its paid Prime service. The Commission argues that Amazon discovered that its customers were accidentally signing up for Prime and were unhappy about it, and that the company nevertheless decided not to fix this confusion because it was making too much money from these accidental signups. To make things worse, Amazon deliberately made it harder to cancel Prime, and celebrated that the new, more complex process resulted in fewer cancellations.
This is historic. Prior to the current administration, the FTC had been in a 40 year decline: underfunded and timid. But the new chair, Lina M. Khan, has brought a muscular, take-no-prisoners approach, working in close coordination with her peers at the DoJ antitrust division and with other agencies to reawaken their long-dormant regulatory powers.
Many of my peers in the tech-critical world were skeptical, or even derisive. No one is accidentally subscribing to Amazon Prime, after all, nor is it especially hard to cancel service. Why not target the New York Times, which only recently — and reluctantly — ended its practice of forcing you to call a human salesperson and endure a lengthy wheedle in order to cancel the same subscription you had initiated online with just three clicks? Why not go after the Wall Street Journal, which still does this?
If tech bloggers were skeptical, the business lobby got downright conspiratorial. FTC Chair Lina Khan, they said, is pursuing her well-known vendetta against poor, defenseless Amazon. That woman has a conflict of interest — after all, she is best known for her scholarship on the problems of Amazon. How dare an antitrust regulator arrive in office with a widely known, deeply felt, and brilliantly analyzed plan for helping the American people get a better deal from a monopoly?
But the reason Khan took after Amazon is that Amazon admitted that it knew it was tricking its customers. What’s more, it put this admission in writing. This is a surprisingly common pathology among tech corporations. They don’t just do crimes — they commit them to writing. It sometimes feels like every tech CEO has a desktop folder called “mens rea” that’s filled with files with names like PREMEDITATED_MURDER_FINAL_FINAL_1.docx.
But Amazon, Meta, Google, and FTX have all done it. The compulsive hubris of tech leaders, their proclivity for communicating by messaging tools and email, and the near-impossibility of deleting all digital trails of these communications have ushered in a new golden age of provable intent. Here’s the crux of the FTC’s complaint, according to a report from Insider: In 2017, Amazon’s Shopper Frustrations teams noticed that a lot of users who signed up for “FREE Two-Day Delivery with Prime” during checkout didn’t realize that in 30 days, they’d be billed $139. These users scrambled to cancel this unexpected charge, and found themselves unable to do so.
So Amazon tinkered with the way it pitched Prime to its customers. They experimented with changing the “Get FREE Two Day Shipping” button to a “Start your 30-day free trial” button, alongside information about the cost of Prime.
In trials, the new button worked — after all, Amazon has excellent UX designers. Signups dropped off a cliff: the company estimated that keeping the new, clearer language would lead to 260,000 fewer Prime subscriptions every year. So, the company dropped it. Despite internal memos documenting how “unintentional sign-ups erode customer trust” and “improvements in clarity during sign-up are needed,” and that the “FREE Two-Day Delivery Pitch” would “tip the balance from customer centricity toward sign-up business goals,” the company killed the experiment.
Another thing the company tried and discarded: an additional confirmation page for any free trial (users had clamored for this, calling the bill at the end of the trial “an unwelcome surprise”). The company also tried — and discarded — a streamlined unsubscribe workflow, one where you simply clicked a button to cancel, and then got a screen telling you you were no longer subscribed, with a button to resubscribe. Not content with confusing 260,000 customers per year into accidentally signing up for Prime, Amazon also made it harder for those accidental Prime subscribers to unsubscribe.
Memos show that Amazon deliberately added confusion and complexity to the account cancellation process. Again, Amazon’s excellent UX designers excelled: cancellations dropped by 14% after the newly complexified cancellation workflow dropped (as one memo put it: “retention appears to be trending positively”).
Amazon called its Prime cancellation process “Iliad,” after Homer’s epic about an endless, grinding war.
All of this is why Amazon is in trouble with the FTC: it knew that its users were being confused into handing over their money on false pretenses, it knew how to end that confusion, and instead of acting to end the confusion, the company deliberately made it worse, and put their intention to do so in writing.
America has a lot of corporate lawlessness, but very little of it ever gets prosecuted. Much of that is down to lack of will, but it’s not just indifference or a lack of resource that lets the corporate corruption rage on unchecked. It’s also because it’s incredibly hard to convict a company of many kinds of corporate wrongdoing, since doing so often requires the government to prove that the company intended to break the law.
Normally, this is a big lift. But not, you know, when the company puts its wicked intentions in writing. Sure, it’s hard to prove that a company knew that its customers were being confused into signing up for a subscription, or that they were struggling with the unsubscription process — but not if the company researches alternatives, concludes that this is true, and then admits to that fact in a memo.
Then it’s not hard at all.
And that’s why Lina Khan is picking on Amazon. Yes, they did something wicked in deliberately setting out to trick their customers into signing up for subscriptions without intending to do so — but more importantly, they did something really stupid: they admitted it.
Others have done the same. In 2012, Mark Zuckerberg sent a middle-of-the-night email to his CFO laying out the case for buying Instagram: Facebook would be acquiring a rival that its own users increasingly preferred. If Facebook acquired the upstart, it could set in motion “network effects” that would make it very hard for future competitors to unseat them. (Meta’s newest product, Threads, enjoyed a massive boost from those network effects when it launched recently.) .”
This is not a thing a CEO should put in writing. When it comes to monopolies, intent matters. If you crush your competition by being better than them, that’s fine. But if you take actions to eliminate competition for its own sake, you’re in hot water.
This doesn’t come up often, because it’s hard to prove that a company pursued some action to attain a monopoly, rather than to improve its offerings or goose its profits. But when the company’s CEO gets out of bed in the middle of the night and sends a fevered email to his CEO saying that’s exactly what he plans on doing, this gets a lot easier to prove.
(Zuck sent a stilted followup email 45 minutes later, explaining — in the cadences of someone who has been sternly lectured by his general counsel — that of course he didn’t mean that Facebook should buy Insta to reduce competition — far from it!)
Another intent-based quagmire is “self-preferencing.” Large platforms like Google can get in big trouble for elevating their own products in search results over superior results from its rivals.
But these prosecutions require proving that Google intended to steer its users towards inferior offerings. If Google sincerely believes that it is showing users the best results — say, if it thinks YouTube videos are better than the ones on Vimeo, and thus elevates YouTube over Vimeo in search listings — then it’s fine.
How on earth can you prove whether Google knowingly put a worse in-house result over a superior third-party offering? It’s a head-scratcher…unless a Googler sends a memo to his boss expressing his frustration at being made to do just that (footnote 192).
Companies don’t merely confess their sins in memos.
Often, they put the confession right in the title. Think of Sam Bankman-Fried and his co-conspirators at FTX, who were all in a group chat called “Wirefraud.”
Amazon does this too. In 2004, the company conceived of a plan to squeeze brutal, “sadistic” discounts out of the small publishers who’d become dependent on Amazon for sales. That plan was called the “Gazelle Project,” because the managers involved were told by CEO Jeff Bezos to “approach these small publishers the way a cheetah would pursue a sickly gazelle” (the only part of this plan that Amazon’s lawyers objected to was the name).
Proving intent is a high bar to hurdle — as the old radio serial tagline goes, “Who knows what evil lurks in the heart of men?”
But large organizations are unwieldy beasts, and they run on paperwork. A busy executive running a many-tentacled oligopolus can’t be expected to keep all their evil plans in their head — much less confine themselves to expressing those plans to trusted Renfields on voice telephone calls, like some kind of cave man. Our tech overlords want to rule the world, and the only way to do so involves a lot of memos.