Enshittification

Based on Wikipedia: Enshittification

The Three-Act Tragedy of Every Platform You've Ever Loved

You remember when Amazon actually helped you find what you wanted to buy. When Facebook showed you posts from your friends instead of ads pretending to be posts from your friends. When Google search results answered your questions on the first page instead of the fifth.

You're not imagining it. Things really did get worse.

In 2022, Canadian writer and digital rights activist Cory Doctorow gave this phenomenon a name that stuck: enshittification. The American Dialect Society crowned it their Word of the Year for 2023. Australia's Macquarie Dictionary did the same for 2024. Both Merriam-Webster and Dictionary.com now list it as a real word.

Enshittification describes a specific pattern of decay. It's not random decline or inevitable entropy. It's a predictable three-act tragedy that plays out on virtually every two-sided platform—meaning any digital service that connects two groups of people, like buyers and sellers, creators and audiences, drivers and riders.

The Three Acts

Here's how the story unfolds, every single time.

Act One: The Courtship. A new platform appears. It's generous to a fault. Amazon sells books below cost. Uber subsidizes rides so heavily that taking a cab feels like highway robbery. Facebook connects you with everyone you've ever known, for free, with no strings attached. The platform is spending money it doesn't have, burning through venture capital like kindling, all to hook you.

And it works. You sign up. Your friends sign up. You start to depend on it.

Act Two: The Pivot. Now that users are locked in—meaning it would be painful or impossible to leave—the platform turns its attention to business customers. Sellers, advertisers, content creators. The platform offers them access to its captive audience. At first, it's a good deal for them too. You can reach millions of customers! You can build a following!

So the businesses sign up. They invest time and money. They become dependent on the platform too.

Act Three: The Extraction. With both sides now locked in, the platform finally reveals its true purpose: extracting maximum value for shareholders. Every knob gets twisted. Every parameter gets adjusted. User experience? Irrelevant. Seller profit margins? Expendable. The platform sits between everyone, taking an ever-larger cut of every transaction, answerable to no one.

And then, eventually, it dies. Or it staggers on as a zombie, too embedded to disappear but too degraded to love.

The Mechanics of Decay

Doctorow has a word for how platforms perform this extraction: twiddling. It means the constant, invisible adjustment of algorithms and rules to squeeze out fractionally more profit. A fraction of a cent more per click. A slightly worse experience that users probably won't notice. One more advertisement disguised as organic content.

Each individual twiddle seems harmless. Collectively, they're corrosive.

The genius—and the tragedy—of enshittification is that by the time users realize what's happening, it's too late to leave. Economists call this switching costs. Your friends are all on Facebook. Your purchase history is on Amazon. Your followers are on Instagram. Walking away means losing everything you've built.

This creates a peculiar kind of monopoly. The platform doesn't just control supply, like a traditional monopoly. It also controls demand. Economists call this second condition a monopsony—a single buyer dominating a market. Platforms like Amazon function as both simultaneously. They're the only game in town whether you're buying or selling.

The Amazon Case Study

Doctorow's original essay focused on Amazon as the prime example of enshittification in action.

In the beginning, Amazon was a miracle. Books cheaper than you could find anywhere else. Free shipping if you paid an annual fee for something called Prime. The site was simple, helpful, designed around helping you find what you wanted.

Then sellers noticed that everyone was shopping on Amazon. So they started selling there too, paying fees for the privilege. Amazon became not just a store but a marketplace.

And then the extraction began.

By 2023, Amazon was taking over 45 percent of the sale price of items sold on its platform. Nearly half. But that's not even the most egregious part. Amazon also turned its search results into a pay-to-play system. Sellers bid against each other for prominent placement, like radio stations used to pay for their songs to get airtime—a practice called payola that was once illegal in broadcasting.

Doctorow reported that if you search for "cat beds" on Amazon, the first five pages of results are about half advertisements. Not the best cat beds. Not the most relevant cat beds. The cat beds whose sellers paid the most to appear there. The search function has been corrupted from a tool that helps you into a tool that helps them.

Facebook: Terminally Enshittified

Doctorow described Facebook as "terminally enshittified," which is about as damning a diagnosis as a platform can receive.

The arc is textbook. Facebook started by connecting friends. That was the whole product. You could see what your actual friends were actually doing. Revolutionary.

Once enough people were using it—once a critical mass made it socially mandatory—Facebook started adding content from media companies into your feed. News articles. Videos. Suddenly publishers depended on Facebook traffic to survive.

Then Facebook changed the rules. Organic reach—meaning the ability to reach people who had specifically chosen to follow you—plummeted. If you wanted your audience to see your content, you had to pay to "boost" it. The algorithm that decided what to show people became deliberately opaque. No one outside Facebook knew exactly how it worked, only that paying money helped.

One particularly grim episode: Facebook told media companies that video was the future. The numbers looked incredible. Publishers pivoted hard, firing writers and hiring video producers. Then it emerged that Facebook had been inflating its video metrics by as much as 900 percent. The statistical fraud led companies to over-invest catastrophically. Some collapsed entirely.

The platform that once connected you to friends now shows you content designed to maximize the time you spend watching ads, selected by an algorithm that even Facebook's own engineers don't fully understand.

Google: From Miracle to Maze

Google Search once felt like magic. You typed a question, you got an answer. Clean. Simple. Accurate.

Google became dominant precisely because its results were so much better than the competition. The company's informal motto was "Don't be evil." The ads were clearly labeled and pushed to the side, never interfering with the actual results you came for.

That era is over.

Today, Google's first page is often dominated by advertisements that look almost identical to real results. An entire industry called Search Engine Optimization, or SEO, has emerged specifically to game Google's algorithm, filling the internet with content designed to rank highly rather than to inform. Some results are outright fraudulent, part of sophisticated scam operations.

Behind the scenes, documents from legal proceedings revealed that Google had been manipulating the advertising market through a program internally nicknamed Jedi Blue—named after the Star Wars Jedi Order, presumably because someone at Google thought market manipulation was heroic.

In January 2023, Google laid off 12,000 employees. Around the same time, the company announced a stock buyback program that, Doctorow calculated, would have been sufficient to pay all of those laid-off workers' salaries for the next 27 years. The priorities were clear.

And now Google is rushing to build AI search tools that won't show you what you asked for, but rather what Google's systems think you should see. The end-to-end principle—the idea that a network should simply deliver what users request—has been abandoned entirely.

Dating Apps: A Perverse Incentive Structure

Dating apps present a particularly twisted case of enshittification, because the product's stated goal directly conflicts with the company's financial interests.

Think about it. A dating app that works too well puts itself out of business. If everyone finds their soulmate in the first week, there's no one left to pay for premium subscriptions.

The optimal outcome for the company is for you to stay single and frustrated—but not so frustrated that you leave. Just frustrated enough to keep paying for another month of premium features. Just hopeful enough to swipe through another hundred profiles. Just lonely enough to believe that the next match might be different.

This creates what economists call a perverse incentive. The better the app works at its stated purpose, the worse the business performs. The worse it works, the more money it makes.

Is it any wonder that so many dating app users report that the services seem to get worse over time?

Uber: The Bait and Switch on Wheels

Uber's story is almost comically straightforward enshittification.

The company started by offering rides that were dramatically cheaper than traditional taxis. This wasn't because Uber had discovered some revolutionary efficiency. It was because venture capitalists were subsidizing every ride, burning billions of dollars to attract users and crush competitors.

Uber also ignored the regulatory systems that governed taxis, like medallions—expensive licenses that limited the number of cabs and theoretically ensured driver quality. This let Uber undercut established players who were following the rules.

Once Uber and its main competitor Lyft had effectively achieved a duopoly—meaning two companies controlling an entire market—the subsidies ended. Prices rose. Driver pay fluctuated wildly through a system called surge pricing, which adjusts rates based on demand.

Riders who had grown accustomed to cheap, convenient transportation were now stuck. Many had abandoned car ownership entirely. The alternatives had been destroyed. All that remained was to pay whatever Uber decided to charge.

Twitter's Accelerated Collapse

When Elon Musk acquired Twitter in late 2022, the platform underwent a case study in accelerated enshittification—decades of decay compressed into months.

The changes came fast. Twitter closed its Application Programming Interface, or API, which is the technical gateway that allowed other software to interact with the platform. This killed third-party apps that many users preferred to Twitter's own. Users who posted their Mastodon handles—Mastodon being a competing social network—found themselves suspended. New restrictions made it impossible to view the site without logging in.

Rate limits capped how many posts you could view per day. A paid subscription service called Twitter Blue, later rebranded as X Premium, introduced verification checkmarks that had previously indicated authentic accounts—but now simply indicated that someone had paid eight dollars a month. Content moderation was dramatically reduced.

The platform that had been the public square of the internet became something unrecognizable within a year.

Reddit's IPO-Driven Decay

In 2023, Reddit announced it would begin charging for API access. This was a direct strike at the ecosystem of third-party apps that many users relied on—apps that were often better designed and more feature-rich than Reddit's official offerings.

The timing was not coincidental. Reddit was preparing for an initial public offering, or IPO—the moment when a company first sells shares to the public. Investors wanted to see a path to profitability. Third-party apps that provided better experiences for users but generated no revenue for Reddit were suddenly a problem that needed solving.

Reddit's CEO Steve Huffman claimed the fee was about stopping artificial intelligence companies from scraping Reddit's data without paying for it. But the practical effect was to make third-party apps financially impossible to operate. Apollo, one of the most popular alternatives, shut down entirely. Hundreds of subreddits went dark in protest. Moderators—the unpaid volunteers who had made Reddit functional for years—organized blackouts against the new policies.

The changes went ahead anyway.

Unity: Enshittification Beyond Social Media

Not all platforms are social networks. Unity is a game engine—software that developers use to build video games. In 2023, Unity announced changes to its licensing model that demonstrated enshittification can happen anywhere two-sided markets exist.

The new terms would have charged developers fees based on how many times their games were installed. Crucially, these terms would apply retroactively to games that had been in development for years, built under different rules. Developers who had invested months or years of work suddenly faced unexpected costs.

The gaming community's response was swift and furious. Many developers announced they would abandon Unity entirely, despite the enormous switching costs involved. The word enshittification was invoked repeatedly. Game designer Sam Barlow used it specifically when explaining why he was leaving the platform.

The exodus was measurable. At the Global Game Jam—an annual event where developers create games in a weekend—Unity usage dropped from 61 percent in 2023 to 36 percent in 2024. Indie developers, who could least afford to be trapped by predatory terms, fled first.

Unity eventually abandoned the most controversial changes. But the trust was broken.

Why Users Stay Anyway

If these platforms have gotten so much worse, why does anyone still use them?

Research suggests two main reasons: fear of missing out, and nostalgia for what the platform used to be.

Users remember when Facebook was good. They remember when Amazon search actually helped them. They keep hoping—irrationally, perhaps—that the old experience might return. Meanwhile, they fear that leaving would mean missing important updates from friends, missing the conversations everyone else is having, missing some essential piece of cultural participation.

So they stay. And they also join new platforms. And they spread their attention across multiple services, none of which fully satisfies them, all of which are at various stages of their own enshittification cycles.

It's exhausting. And it's by design.

The Harm Goes Deeper Than Annoyance

One study by researchers Ardoline and Lenzo found that platform decay causes genuine cognitive and moral harm. When a platform that once helped you find information starts actively making that harder, your ability to process information degrades. You become less capable of distinguishing reliable sources from unreliable ones. The constant manipulation affects how you think.

Other researchers have applied the concept of enshittification to labor markets. In a 2025 study, scholars Maffie and Hurtado argued that gig economy platforms follow the same pattern—first offering workers favorable conditions to attract them, then steadily degrading those conditions once workers are dependent on the platform. The result is increased precarity, opacity, and power imbalances.

Enshittification, in other words, isn't just about consumer inconvenience. It's reshaping labor markets and possibly affecting how millions of people think.

Two Paths Out

Doctorow doesn't just diagnose the problem. He proposes solutions, focused on two principles.

The first is the end-to-end principle. This is an old idea from network engineering that says the role of a network is to reliably deliver data from senders to receivers. Nothing more. The network shouldn't decide what you should see or in what order. It should just deliver what you asked for.

Applied to platforms, this would mean that if you subscribe to a content creator, you see everything they post. No algorithm deciding what to hide. Search engines would show exact matches for your queries before sponsored results, not after. The platform would be a pipe, not a gatekeeper.

The second principle is the right of exit. If you're unhappy with a platform, you should be able to leave without losing everything. This requires interoperability—the ability for different services to work together. If you could take your social graph with you when you left Facebook, switching costs would plummet. If you could export the movies you purchased on one platform and watch them on another, digital rights management would lose its lock-in power.

Both principles are technically feasible. Neither is likely to happen without regulatory intervention, because platforms have no incentive to make leaving easy. Their entire business model depends on making leaving hard.

The Enshittocene

In a 2024 Financial Times piece, Doctorow went further. He argued that enshittification is "coming for absolutely everything." Not just tech platforms, but the broader structure of how business operates in a digitally mediated world.

He coined a term for this era: the enshittocene. Like the Anthropocene—the geological epoch defined by human impact on Earth—the enshittocene describes an age defined by the systematic degradation of digital services for short-term profit.

Henry Farrell, a political scientist, has applied the concept even more broadly—to American power itself. Military hardware. The dollar as global reserve currency. Satellite constellations. All of it, Farrell suggests, follows similar patterns of initial value creation followed by extractive decline.

Maybe that's too expansive. Maybe enshittification is specifically a phenomenon of two-sided digital markets and shouldn't be stretched to describe everything that gets worse over time.

But the core insight feels undeniable: platforms that position themselves as neutral intermediaries inevitably face the temptation to exploit that position. And without countervailing forces—competition, regulation, interoperability—that temptation becomes irresistible.

What Comes Next

In October 2025, Doctorow released a book titled Enshittification: Why Everything Suddenly Got Worse and What To Do About It. The title suggests he believes there are paths forward, ways to reverse the decay or at least stop its spread.

Whether those solutions can be implemented at scale remains to be seen. The forces driving enshittification—shareholder primacy, network effects, regulatory capture—are enormously powerful. Platforms that have already extracted themselves from competition have little reason to change.

But naming a problem is the first step toward solving it. Before Doctorow's essay, the pattern was something people sensed but couldn't quite articulate. Now there's a word for it—an admittedly vulgar word, but one that captures the visceral wrongness of watching something useful become something predatory.

The platforms got worse. You noticed. You were right to be angry.

Now the question is what we do about it.